AD-AIO 370 FEDERAL AVIATION ADMINISTRATION WASHINGTON C F/6 17/7

UNLSIID THE FAA PLANS AND PROGRAMS FOR THE FUTURE AIRPORT AND AIR TRAFF-ETC(U)

20 17ffllfllfflffllf

UCL umSIFIEDuu1E-2EEEEE

mom-C

IE FAA PLANS AND fROGRAMSFOR THE FUTURE

o AIRPORT AND AIR TRAFFIC CONTROL SYSTEM.

Preseintatios for theOffice if Technology Assessment Smiar

" Prepared by

I. DEPARTMENT OF TRANSPUATISNfEL AVIATION ADIUIISTUUIH

Wsblagtml, I. C. 2N10

81 6 18 008Oise"A

Technical kepor- O cwrnentwon lf~

~. T,,i. 5 R.Po-0t

The FAA Plans and Progrars for the Future November 19F'r

Airport and Air Traffic Contrel System 6P,~~~Cg~

-____ ____ __ 8. P or-.g Cg . .o R.por N7 A.,o ,,~

9 PI.r ri~O;o Ada-%,e~. 10 Rd,. U-, h..~ A

Federal Aviation Administratio'n ______________

S.9C Independence Ave., S. 1.Co'cma

Washington, D.C. 20591 ___________ ___

12 lyp 0# R.p.- -4 .. e a..12 Si-oloi.g A9*ncn Nor. d-3 Adde,-

Deoartment of TransportationFederal Aviatir~r Administration______ _______

Wash inaton, D.C. 2959l pn1.

15 Suppl.vneion Nit

16 Abotroc'

tY-C'Uch the NZ1,1cra"

Fr~n D 0T F 170.~~ m6'i'~0i m?'

TABLE OF CONTENTS,

Page

Introduction--FAA's Role in Providing a Safe and IEfficient System - Mr. Neal A. Blake, Deputy AssociateAdministrator, Engineering and Development

. FAA Forecasts of Aviation Activity, 1981-1992 - 4Airport and Airway System Capacity and Delay Overview*, 11Mr. Harvey B. Safeer, Director, Office of Aviation Policyand Plans

.Today's ATC System-Problems and Need for Change * Mr. Raynond 38VanVuren, Director, Air Traffic Service

An Overview of the FAA Engineering and Development Programs 53Mr. Robert W. Wedan, Director, Systems Research andDevelopment Service

Scenario for the Future System - 88The Roadmap to the System of the Future - 94The Impact of Alternative Approaches in Air Traffic 106Contol System Evolution , Mr. Siegbert B. Poritzky, Director,Office of Systems Engineering Management

.Airport Capacity Increases--Opportunities, Limitations, and Choicese 109Mr. Siegbert B. Poritzky, Director, Office of Systems EngineeringManagement

- Near Term System Improvements --Mr. Martin T. Pozesky, Deputy 132Director, Systems Research and-Development Service

Long Term ATC System Improvements * Dr. Edmund J. Koenke, 140Acting Deputy Director, Office of Systems EngineeringManagement

Nontechnological Alternatives for Balancing Airport/Airspace 152Supply and Demand !Mr. Harvey B. Safee ,ir-ect.o O,_f cof Aviation Policy and Plans. ' j

__U r

p- •' . -f++ +++ }

INTRODUCTION -- FAA'S ROLE IN PROVIDING

A SAFE AND EFFICIENT SYSTEM

byNeal A. Blake

Deputy Associate Administratorfor Engineering and Development

I am pleased to be here this morning to introduce a very full day of briefingsat this OTA Seminar on the Airport and Air Traffic Control System. We willattempt to respond specifically to the wishes of the Advisory Panel and theOTA staff in presenting the information you asked for.

0 We will begin with two briefings on "Traffic Forecasts on AviationActivity, 1981-1982" and "Airport and Airway System Capacity atid DelayOverview," to be given by Mr. Harvey Safeer, Director, Office ofAviation Policy and Plans.

o Next will be a description of "Today's ATC System--Problems and Needfor Change," by Mr. Raymond Van Vuren, Director, Air Traffic Service.

o Next will be briefings on the system of tomorrow, including "AnOverview of the FAA Engineering and Development Program," to bepresented by Mr. Robert Wedan, Director, Systems Research andDevelopment Service. Then follows a description of the "Scenario forthe Future System," "The Roadmap to the System of the Future," and "TheImpact of Alternative Approaches in Air Traffic Control SystemEvolution," to be given by Mr. Siegbert Poritzky, Director, Office ofSystems Engineering Management.

o After lunch, Mr. Poritzky will continue with a briefing on bottom linefindings from FAA's extensive airport capacity and delay activities.

0 Mr. Martin Pozesky, Deputy Director, Systems Research and DevelopmentService, will describe several of our near-term research anddevelopment activities for system improvement.

0 He will be followed by a briefing on several of FAA's longer termsystem improvement activities, by Dr. Edmund Koenke, Acting DeputyDirector, Office of Systems Engineering Management.

o Mr. Safeer will then return to discuss "Non-Technological Alternatives

for Balancing Airport/Airspace Supply and Demand," and

0 I will close the formal presentations with a brief wrap-up.

Your staff has laid on a very extensive program, and we will need your help toget through it quickly and effectively.

1,

Before we start, I would like to spend a moment or two discussing FAA's 'olein providing a safe and efficient system, although I know that each of youhere is very much familiar with FAA's roles and activities.

The basic tasks that FAA performs are:

o The management and operation of the National Airspace System;

o The issuance and enforcement of safety rules and regulations;

0 The certification of airmen, aircraft, aircraft components, airagencies and airports; and

o The conduct of aviation safety-related research and development.

The dimensions of the system are large. At the end of Calendar Year 1979there were some 800,000 active FAA certificated pilots, including more than200,000 student pilots. Among non-pilot airmen with FAA certificates are229,000 mechanics, nearly 26,000 control tower operators, and about 33,000flight engineers. There are some 2,660 air carrier aircraft; 202,000 generalaviation aircraft; some 2,500 pilot schools; 134 mechanic schools; 2,750repair stations in the United States, and 107 overseas.

FAA employs upwards of 56,000 people, including over 16,000 air trafficcontrollers. There are 431 air traffic control towers, 25 control centers,and 319 Flight Service Stations. On the equipment side, there are about10,000 navigational aids operated to define the 351,000 miles of the FederalAirway System. The airport network consists of nearly 15,000 landing fields,including 4,761 publicly-owned airports, of which 736 serve both scheduledcarriers and general aviation aircraft.

Mr. Safeer will talk to you in some detail about our forecasts. He will showthat the U.S. aviation industry continues to grow on almost all fronts. Yet,there are problems ahead. A few days ago Administrator Bond said:

"...Despite the apparent strength of our industry and the outlook forbusiness ahead, aviation is at a crossroad that threatens disaster,born perhaps because of the increasing acceptance and dependence ofpeople everywhere on air transportation, and fueled, certainly, byincreasing economic pressures on industry and government. Thewell-being of the entire industry is threatened."

"The...aviation community has become unhealthily fractious. In pursuit

of narrow interests, too many of us ignore the general health ofaviation. Unless we can get together, and soon, on the issuesconfronting us; unless we manifest a working belief that ourprofessional intentions are truly mutual--one with another's--we maysoon find ourselves squarely in the midst of irrevocable restraintsthat may irrepairably constrain aviation's growth."

Laws.

In briefing you today, you will see that we are working on a number ofimprovements and remedies. You will also see that there is a clearrequirement for major capital investment in the system during this and thenext decade. The funding levels for these investments which FAA has proposedrepresent a growth budget. At the same time, we all know that we areoperating in an environment of fiscal restraints at all levels of government.The critical question is whether the Federal budget for aviation capitalinvestment can grow Fast enough to accommodate the dimensions we foresee.

A question undoubtedly before you is whether we know where we are going andwhat we are doing to meet the need. Messrs. Van Vuren and Safeer will tellyou of the current system, the problems in it, and. the economic and policyissues confronting the community. We will describe to you something of theEngineering and Development program intended to lead us to the Year 2000 in arational and progressive fashion.

Much of the aviation community is aware of what we are doing through long-timeexposure and exchanges of view which occur on an almost continuous basis. Theeffort we called "New Engineering and Development Initiatives--Policy andTechnology Choices" was an attempt to lay out before the entire community whatwe saw as problems and solutions. We asked for and received guidance on manyof the same issues you are confronting. The people involved in this processspent literally months bringing themselves up-to-date on what we were doing,and developing thoughtful recommendations which we are taking seriouslyindeed. In the bottom line, the New E&D Initiatives results represent strongand informed support for our Engineering and Development program andrecommendations for strengthening it in a number of areas. We are trying todo so within the very real limitations on available resources.

I mentioned earlier that there would be a discussion of alternative paths tothe system of tomorrow, and questions about those alternative paths havealready been raised in the staff assessments as they were during the NewEngineering and Development Initiatives effort and often before. We are wellaware that there are alternatives and have been studying a number of them.Such issues as the balance of responsibility between the pilot and the controlsystem, the future air traffic control surveillance system, the number of airtraffic control centers, the future navigation system, the acceptable levelsof automatioii, and others, all represent forks in the road on which decisionswill need to be made. But the decisions will have to be made in such a waythat one unresolved issue doesn't hold up the train. We will be discussingthese issues in more detail later in our presentations.

The briefings in this seminar will give you an idea of the scope of our workand will, we hone, place you in a better position to make meaningful andthoughtful assessments of that need.

FAA FORECASTS OF AVIATION ACTIVITY 1981 - 1992

byHarvey B. Safeer

Director, Office of AviationPolicy and Plans

Forecasting is still as much an art as it is a science. There are no modelsor sets of models, however, sophisticated, that can "predict" futuresocioeconomic events. Most models of socioeconomic events merely state whatwe expect to happen if certain other events take place, the so-calledexogenous variables, and if the relationship between the exogenous variablesand what we are trying to forecast is as we have postulated it to be.

Forecasts of aviation activity are based upon other forecasts of generaleconomic activity and relationships between what we expect to happen in thegeneral economy and how these events will affect aviation. Thus, it isimportant to discuss not only the results of our forecasting efforts, butalso the assumptions which we have either accepted, based upon the work ofothers, or those which we have either accepted, based upon the work ofothers, or those which we have made ourselves.

The key structural assumption which FAA has made with respect to the airtransportation industry is that the basic relationship between the FederalGovernment and the industry will continue to be one of economic deregulation.

Based upon the Wharton long-term industry and economic forecasting model,FAA is using the following economic assumptions for the period 1980-1992:

1. Real gross national product is forecast to grow at an annual

compound rate of 2.7 percent;

2. Employment is expected to grow at an annual compound rate of 1.3percent;

3. Consumer price index is expected to rise some 11.7 percent in 1981,but by 1992, the rate of increase is expected to slowdown to 7.3percent, with a compound annual rate of growth of 8.2 percent;

4. Real disposable personal income is expected to grow at a compoundannual rate of 2.8 percent;

5. Fuel prices, based on the Wharton projection of the oil and gasdeflation, are forecast to grow by 225 percent between 1980 and1992. The forecast assumes, however, that fuel will be availablefor aviation; and

6. The unemployment rate is forecast to peak in 1981, and then declineto 5.0 percent by 1992.

4i

In addition to these general economic assumptions, FAA has made a series ofassumptions specific to aviation:

1. General aviation fuel costs will increase at an average annual rateof 10.4 percent;

2. The average annual fixed cost of owning and operating a general

aviation aircraft will increase at an annual rate of about 6 percent;

3. The overall certificated air carrier average passenger trip length

is expected to grow at the historical rate of 3 miles per year; and

4. Average seats per aircraft are expected to increase about 4 seatsper year.

Two additional assumptions are: Load factors are expected to increase from

about 61 percent in 1980 to 63 percent in 1984 and beyond, and revenue per

passenger mile will increase in current dollars about 5 percent per year,

but will decrease in constant dollars from the current 4.5 cents to 3.7

cents.

A change in any of these assumptions will, of course, affect the forecasts.

Thus, FAA has continued the practice of developing a set of forecasts based

upon alternative future scenarios. When you read the descriptions of these

alternative scenarios, don't be put off by the postulated events that lead

to the alternative assumptions. Just remember that what seemed to be fat

out and improbable to some in the early 1960's--deregulation, high fuel

prices--are facts of life today. The important use of these scenarios is to

postulate the possible changes in the exogenous variables and, in turn, to

see how these changes impact our forecast of aviation activity.

The key forecasts which are expected to affect FAA policy and investmentdecision which must be made in the next few years are summarized below.

The air carrier industry is still undergoing a period of adjustment, not

only to deregulation but also to higher fuel prices, environmental

regulations, changing relationships within the industry and the introduction

of new equipment. The jet age is slightly more than 20 years old an we willbe seeing the third generation of new aircraft entering the fleet.

General aviation is also experiencing more subtle, but nevertheless critical

change. In addition to having to cope with higher fuel prices and other

costs, general aviation is being called upon to serve an increasing role in

providing transportation which is essential to economic growth and

development.

SI

Domestic air carrier revenue passenger enplanements are expected to resumetheir growth in 1981 concurrent with recovery from the current recession.Over the 12-year period, we expect domestic revenue passenger enplanementsto grow by an average 4.3 percent per year, while revenue passenger milesare expected to grow by some 4.8 percent per year.

Commuter carrier are expected to sustain a higher average annual growthrate, particularly as new equipment enters the fleet over the next few years.

We expect the general aviation fleet and total hours flown to increase atmodest rates over the next 12 years. These growth rates, however, tend tomask the expected growth in the use of higher performance, better equipped,multi-engine aircraft which are entering the fleet at a rate which is doublethat of single engine aircraft.

Given these general forecasts of aviation activity, we expect a moderaterate of growth in FAA workload over the next 12 years, on the order of 3percent per year for tower and center activities, approximately 4 percentfor flight service station activities.

Total operations at airports with FAA traffic control service are forecastto increase by 43 percent between 1980 and 1992. However, this expectedincrease is only part of the story. We expect to see a continuation of thetrend toward increased participation in the system of air taxis (includingcommuters) and general aviation itinerant flying. Thus, air carrieroperations are expected to increase by only 21 percent over this time periodwhile air taxi and general aviation itinerant operations are expected togrow by 98 percent and 49 percent respectively. General aviation localoperations are forecast to grow by 39 percent.

The net effect of these differential growth rates is a redistrioution in themix of operations in the system. This shift in the mix of aircraft typesusing towered airports has its implications for the operation of the airtraffic, the greater the problems associated with local flow controlmanagement. See Figure 1.

Instrument operations at towered airports are expected to increase at aslightly faster rate than total operations. This represents a continuationof the trend toward more shophisticated equipage of general aviationaircraft and their increased use for business and commercial purposes, aswell as the effect of additional TCA's and TRSA's.

This trend is also reflected in the number of TFR aircraft handled by ourair route traffic control centers. While total activity is expected toincrease by 46 percent, air carrier aircraft handled are expected toincrease by only 22 percent. Air taxi (commuter) and general aviationaircraft handled are expected to increase by 124 percent and 86 percentrespectively. Once again, these differential growth rates will result in aredistribution of the relative share of the workload. By 1992, the centerswill be handling almost as many general aviation aircraft as air carripraircraft. See Figure 2. 0t

4

As I indicated earlier, these forecasts are hased upon a set of assumptionsand forecasts of general economic activity. We also generated a series offorecasts based upon alternative sets of assumptions. While you can readthe details of the scenarios and the resultant forecasts in the report, Ithink that it is important to focus on the implications of these alternativeforecasts for 1992. Two of the scenarios, "Economic Expansion" and "EnergyConservation" tend to bracket the baseline forecast. In fact, if you lookat the numbers carefully, you will observe that the general trend of thebaseline forecast is closer to the "Energy Conservation" scenario. Thegeneral trend for the baseline and these first two alternative scenarios, isowever, economic growth and concurrent aviation growth, albeit at differingrates. It is only under the third alternative scenario, "Stagflation,"where we see a significant departure from a growth trend. See Figure 3.

The implications are significant. If we truly believe that our economy isgoing to grow at all over the next decade, we must accept the logicalextension of that belief, which is that aviation will grow. To not acceptthe strong interdependence between the economy and aviation's future is toassume that the very structure of our air transportation system will changeover the next 12 years. I do not foresee any technological, social, oreconomic changes which will be strong enough, in and of themselves, toeither change these relationships signficantly or reverse the long-termtrends which we are forecasting. There may be cyclical perturbations aboutthis trend, such as the recessions of 1975 and 1980, but so long as wecontinue to provide an adequate infrastrucutre for the air transportationsystem, the trend for all types of aviation activity is growth.

In conclusion, let me restate my initial premise: forecasting is at best aninexact science. Over the long-term, we can generate probable trends andidentify both the forces underlying those trends and the forces which cancause deviations. If we can agree that the trends have been correctlyidentified, then we have developed a mutual framework for future planningand policy development.

CN

C* O00

0 0 0 00LO CY

(NN

X\0

6I-wo

000 0 0(V C) I-.

P9IPUeH OIlleJi U:HI 10 IU93J~d

9

0

11 LL ) 00

-. 0C',

0 C CA 00 LO a~ )

~ 0

a)

a~ ) LO 1 Q

C4u

0

Cu 0 c ) coo ~ CLo V

mC 0$ 06 6 L6

00

C

o 0. -

0.0

0

o 0 U-

AIRPORT AND AIRWAY SYSTEM fUAPACITYAND DELAY OVERVTEW

byHarvey B. Safeer

Director of Aviation Policy and Plans

I. CONCEPTS OF CAPACITY AND DELAY

Capacity and delay are illusive concepts, surrounded by confusion andmisunderstanding. A substantial part of this problem is the resultof multiple definitions being interchangeably used, and incompletedata being collected by multiple sources for varying purposes. Theproblem is further compounded by the difficulty in determining causeand effect relationships from the data which are available. Thefollowing discussion is intended to set the framework within whichone can discuss the dual issues of capacity and delay.

A. CAPACITY

Two principle definitions of capacity have been advanced indiscussions of terminal area capacity: (1) a so-called"practical" measure, and (2) a "throughput" measure(References 15, 20). The "practical" measure provides a measureof capacity which is defined with respect to a maximumacceptable average delay. (PANCAP is one well-known measure ofthis type.) The "throughput" measure is a measure of capacityindependent of delay; it assumes that an aircraft will always bepresent waiting to use the terminal. A clear distinctionbetween the two requires a brief description of the delayprocess.

If all users of a system consistently arrived at evenly spacedintervals, the system could provide service hourly to a numberof users equal to the service time in minutes divided into 60.This is the maximum possible service rate and is the"throughput" measure of capacity. Unfortunately, system usersdo not arrive consistently at evenly spaced intervals.Sometimes several users arrive at one time and sometimes no onearrives. As a consequence, some of those who arrive at the sametime as do others must be delayed. And at times no one is usingthe system. The "practical" capacity measure is the number ofusers that can be served hourly with the average user incurringdelay of a certain level, after taking into account theunevenness of arrivals.

The two measures are illustrated by Figure 1 which indicates thetheoretical relationship between capacity and delay. As can beseen, the "throughput" measure is the maximum capacity

II

attainable. It results in very high average delaylevels--infinite at the limit--as a consequence of theunevenness of arrivals. The "practical" measure is less thanthe "throughput" measure. It is that level of capacity thatcorresponds to a given acceptable level of delay.

Although both measures have been used in studies of terminaldelay, the "throughput" measure seems to have received moreattention in later work (Reference 20). This is because it isrelatively simple to calculate and independent of delay. Inaddition, being independent of delay, it is not affected by andwill not vary with different delay calculation schemes. The"throughput" measure is thus comparable from situation tosituation, regardless of the delay estimation techniquesemployed in each situation.

It should be pointed out that the relationship depicted inFigure 1 may not be directly observable in the real world.Figure I is drawn on the presumption of a single processing ratefor all levels of operations. In reality, the processing rateoften varies directly with the number of operations for a numberof reasons. (See Section III for a more detailed discussion ofthe factors affecting the delivery rate.) Staffing levels arealmost always positively correlated with expected traffic.Controller productivity may also increase as demand increases.And some systems (such as the en route airway system) may havemore than one processing system (route between two terminals),each with a different processing time. As a waiting 'inedevelops behind the most efficient system, some of those waitingmay turn to the second, third, and so on, most efficientsystem. Users served by these less efficient systems, whileactually spending more time being served, will save enough timewaiting for service to reduce overall time.

The impact of the processing rate increasing as the level ofoperation does will be to shift the delay-capacity relationshipsdownward. The observed relationship will frequently be belowthe curve as drawn in Figure 1. And, if the processing rateshould increase fast enough over a particular range ofoperations, the observed level of delay might actually declineover a particular range of operations.

B. DELAY

1. Acceptable Delay

Strictly speaking, delay will occur each time an aircraftis required to utilize other than the optimum route betweentwo terminals. Whether or not this delay is significant,however, depends on that level of delay which is judged to

1I1

FIGURE 1RELATIONSHIP BETWEEN CAPACITY

AND AVERAGE DELAY

AVE RAGEDELAY

maximum acceptable delay

practical through putcapacity capacity

NUMBER OF OPERATIONS

be "acceptable." "Acceptable" delay is, thus, a standard of theefficiency with which the en route system is expected tooperate. Delays in excess of this level are an indication ofsubstandard performance and are a signal that additional systemexpansion is required.

Adoption of "acceptable" delay standards is an exercise inpublic policy and is ultimately a political decision.Nonetheless, there are several criteria which thepolicymaker should consider in the establishment of thesestandards, including the following. First, part of alldelay occurs because of conditions beyond anyone'scontrol. Such conditions include variations in wind,weather, pilot proficiency and aircraft performance.Because there is little that can be done about suchfactors, there is little choice but to treat the delay theycause as "acceptable." Second, the economics of delayreduction investments should be considered. Under a stricteconomic criterion, investments in delay reduction shouldcontinue to be made until the benefits associated with suchinvestments just equal the cost of undertaking them. Thelevel of "acceptable" delay is that level which prevailswhen this economic condition obtains. "Acceptable" delayis, thus, that level of delay which it does not pay toeliminate. Third, the air traveler's ability to perceivesmall segments of time--say 30 seconds--might beconsidered. If it could be shown that air travelers arenot aware of delays of 1/2, 1, 2, 3, or 4 minutes, it wouldbe hard to justify investment expenditures to eliminatesuch delays to passengers. Fourth, it must be recognizedthat delay is a random phenomenon. Sometimes a flightbetween two terminals will experience small or no delays,while at other times delays will be large. This willgenerate problems in terms of scheduling, passengerconnections, and maximum aircraft flying times.Accordingly, the policymaker must consider the maximumacceptable delay which, when encountered, would undulydisrupt the air transportation system.

2. Delay Classifications

Delay is commonly classified by the segment of airspacewith which it is associated. This leads to confusion as towhere aircraft actually experience delay and as to where

the events that cause the delay occur. Informationconcerning the airspace segment where the factors whichcause delay occur is important in that it focuses attentionon segments of airspace with insufficient capacity.Knowledge of where the delays actually are experienced isimportant in that it identifies where the delayed aircraft

14

must actually be accommodated. Moreov,, since some agencydelay programs such as "flow control" seek to move delays f-omone air route segment to another, such information is essentialif these programs are to be evaluated.

Figure 2 presents a matrix of delay classifications whichindicates where delay originates and where it actuallyoccurs. Airspace segmeots where delay originates arelisted across the top. Airspace segments where delaysactually occur are listed in the left margin. Each box isassigned a Roman Numeral-Letter designation and representsa different delay classification. The principal diagonalof the matrix--enclosed in the solid line--representsdelays which occur in the same airspace segments as doesits cause. Those boxes which are above the diagonalrepresent delays which take place in segments before theone in which the delay is caused. The shaded , 2a whichlies below the principal diagonal does not requireclassification.

FIGURE 2

DELAY CLASSIFICATIONS

Location Airspace Where Delay CausecDelay Departure-- -En Route En Route Arrival--

Experienced Terminal (CONUS) (Oceanic) TerminalM1 TI)( l)iV)

A. Departure _7

Terminal IA IIA IlIA IVA

B. En Route lIB IIIB IVB

C. Arrival ,7/Terminal .. / IVC

Delay caused in a particular airspace segment cannot

actually take place in airspace segments which the aircraft

encounters after the segment of delay origin. (As an

analogy, water backs up behind a dam, not in f-ont of it.)An exception might be when departure delays cause arrivaldelays because there are too many aircraft on the airportsurface to permit additional aircraft to be landed.Although these types of exception do occur, they are forthe most part atypical. The following paragraphs describeeach type of delay and where it occurs.

a. Departure-Terminal: This delay (IA) is caused byevents at the departure terminal and occursexclusively at this terminal. The most frequent causeis weather. Because the situation is known to allpotential departures, this type of delay is takenalmost exclusively on the ground--where it is leastcostly.

b. En Route (CONUS): En route delay occurs whenever anaircraft must take longer to complete a trip betweentwo terminal areas than the minimum achievable time.Such delay occurs because the optimum route is notavailable for the aircraft for one of a number ofreasons: (1) traffic volume between the two terminalareas may exceed that which may be accommodated by thfoptimum route, (2) severe weather may result in theoptimum route being closed, (3) heavy traffic volumeacross the optimum route may require that an alternateroute be flown.

Delays generated by en route events most likely willoccur in the en route airspace (iB). It is possibleunder extreme conditions that such delays may back upinto the terminal area (IIA). If they do back up intothe departure terminal, they will most likely be takenon the ground.

c. En Route (Oceanic): En route oceanic delay, like enroute CONUS delay, occurs when an aircraft takeslonger to complete a trip than the minimumachievable. It is caused by the same factors asdomestic delay. In the North Atlantic, limitedoptimum or near optimum air routes relative to demandfor 3ervice are likely to be the primary cause. Thesedelays may be experienced en route (111B) or he pushedback to the departure terminal area (IliA) where theyusually will be taken on the ground.

d. Arrival Terminal: Delays generated in the arrivalterminal airspace occur because the terminal cannotland aircraft at the rate they are arriving. Thisdelay may actually occur in the terminal area (IVC)

I',

but most often backs up into er route airspace (VB'so as to avoid congestion in the terminal a'ea andpermit aircraft to hold at higher altitudes where the),are more fuel efficient. 'Note that most holdingstacks are in en route airspace.) At times, thesedelays may back ip all the way to the departureterminal where aircraft bound for congested terminalswill be held on the ground (IVA).

I. DELAY MEASUREMENT

There are four currently or potentially available sources of delaydata.

A. NATIONAL AIRSPACE COMMANC C(ENTER 1NASCOM)

About sixty airports report cn a daily basis their delays of30 minutes or longer. Thtcse data are received at NASCOM andmaintained by the Air Traffic Service. The data include abeginning and ending time for each series of delays, the numberof delays during that period, and a primary and secondary causeof delays for that period. The determination of a NASCOM delayis, in practice, a subjective decision of the controller. Thequality of reporting is subject to the variation in controllerworkload.

These data are readily available in a computer data base andprovide a very broad view of serious delay problems. Their lackof precision limits their use in analyzing delay causes, butthey may provide an immediate ability to monitor delay trends ata large number of airports.

B. PERFORMANCE MEASUREMENT SYSTEM (PMS)

The Air Traffic Service also maintains, but not o-i computer,records of delays received through the PMS. These delays areofficially described as being 15 minutes or longer but inpractice shorter delays may be included. The definition andreporting of PMS delays are subject to the same constraints as

NASCOM delays. The number of delays and airport conditions arereported by hour by about twenty airports.

C. STANDARD AIR CARRIER DELAY REPORTING SYSTEM

Eastern Air Lines, American Airlines, and Vnited Air Linesreport delay data on about thirty airports'to the Office ofAviation Policy. These are delays sorted by phase of flight andare in a computer data base. The causes of delays are notincluded, but the data provide a relativelr detailed means of

monitoring delay trends. The definition of a delay is based ona nominal standard for ground time and on computer-projectedflight time.

\I

D. OSEM SYSTEM

The Office of Systems Engineering Management developed a methodof monitoring delay trends at airports, using CAB data onoperational times actually experienced by air carrier flights.These data provide monthly estimates of the flight times betweenmajor airports. An arbitrary standard flight time must besubtracted to establish estimates of delays, but the results canbe used to detect trends in delays.

Of the readily available sources of delay data, only the air carrierreporting system employs a standard of minimum flight time andsystematically reports deviations from the standard.

III. CAUSES OF AIRSPACE/AIRPORT SYSTEM DELAYS

FAA studies of delay at specific airports have identified six genericcauses of airspace/airport system delay:

o The proximity of other airports;o Air traffic control rules, regulations, and proceIjres;o Physical properties of the airspace/airfield;o Meteorology;o Operational procedures; ando Aircraft operating demand.

A. THE PROXIMITY OF OTHER AIRPORTS

The proximity of other airports to the specific airport beinganalyzed affects delays to the extent that their operationslimit the paths over which aircraft may be vectored to or fromthe subject airport, or must be coordinated through approachcontrol or the tower. Delays can be the result of a requirementto hold departures at one airport until arrivals have cleared atthe other one, or a gap may be required in the arrival streamfor one airport to accommodate arrivals to or departures fromthe other airport.

Delays may also be incurred when less than optimal routing isrequired in order to preclude incursion into the airspace of anadjacent airport. These routings can take the form of longerdistances before turns are initiated in order to attainsufficient altitude to climb over conflicting approach paths orlong approach legs at low altitudes to pass under conflictingflight tracks.

B. AIR TRAFFIC CONTROL (ATC) RULES, REGULATIONS, AND PROCEDURES

Although designed to ensure operational safety in the airportenvironment, certain ATC rules, regulations, and procedureslimit, to some extent, total system capacities achievable and

affect delays. While ATC rLs ;rd :,qb , CrPar absolutelynecessary for safety of operation, their relationship tocapacity and delay should be understood. The rules andregulations most affecting capacity and dolay are thoseregarding separation requirements between ar-iving and departingaircraft. While it is nct suggested that delay reduction beachieved through modifying the rules or procedures, one shouldunderstand why a certain level of delay is inherent any time youhave a heterogeneous mix of aircraft operating at an airport.

1. Arrival Separations

At present, current ATC rules under !FR conditionsstipulate that certain distances must be maintained betweenarriving aircraft of different weiqht classes. The currentIFR separation standards are 3, q, 5. ? , 6 nalJtica1 miles(LL, HH, H, LS. HS)*. in com ' arison, the separation underVFR is significantly less under stiiratoed trafficconditions.

2. Runway Occupancy

The second basic ATC rule is that two aircraft may not bothoccupy the same runway. Once the first aircraft crossesthe threshold, it has sole possession of the runway untilit exits. The second aircraft must be spaced such that itdoes not cross the runway threshold until the first hascleared the runway.

3. Departure/Arrival Spacing

Current operating rules prohibit the initiation of adeparture unless the followinq arrival is more thantwo miles out from the thresholj.

4. Departure Separation

Current IFR operatinq rules define the minimum drpartingseparation. Further, due to the wake vortex problem, VFR

*5, L, H refer to ATC weight classes:

Small (S): Less than 12,500#Large (L): Between 12,500# and 300,000#Heavy (H): Greater than 300,000#

Notation "HL" denotes heavy followed by a large aircraft..The notation "LL" includes all pairinqs not otherwispspecified (i.e., SS, SL, SH, U.. [H.

I.

standards for aircraft following a heavy are the same asIFR standards to ensure the safety of aircraft whichtakeoff after a heavy aircraft. The current IFR standardsare HH: 90 seconds; HL, HS: 120 seconds; all others:60 seconds.

C. PHYSICAL PROPERTIES OF THE AIRSPACE/AIRFIELD

The physical properties of an airport's airspace/airfielddetermine not only the ability of the entire system toaccommodate various aircraft types, but also the operatingefficiency, in terms of capacity and delay, of theconfigurations in which the airspace/airfield functions. Thefollowing are examples of physical properties which couldinfluence delays:

o Obstructions;o Displaced thresholds reducing usable runway length;o Shoulders on runways;o Intersection and exit locations 3nd number;o Location of airline gates vis-a-vis runway exits; ando Weight limitations on runway segments.

The significance of these physical limitations will vary fromairport to airport.

D. METEOROLOGICAL CONDITIONS

The operational strategy of an airfield is governed to a largeextent by considerations of ceiling, visibility, precipitation,and prevailing wind directions. These conditions determine notonly what runway configuration will be in operation, but thecontrol procedures to be used in processing aircraft to and fromthe field. Figure 3 shows 18 possible runway use combinationsat O'Hare International Airport. An arrowhead pointing to arunway end indicates landing directions, an arrowhead emanatingfrom the runway end indicates takeoff direction. Figure 4 showsthe combined effects of weather (IFR versus VFR) and runwayconfiguration specific capacity on average delay per operation.With a constant demand, average delay can range between3 minutes per operation and 37 minutes per operation.Therefore, when the winds dictate the use of a high delayconfiguration, a premium in terms of increased delay, is paidfor its use.

Ceiling and visibility also affect the selection of operatingconfigurations. Depending upon instrumentation and conditionsaffecting their use, landing minimums can vary from runway torunway, necessitating adjusting the operating configuration tothe prevailing ceiling and visibility conditions irrespective of

.mmIm()J

RUNMAY CuNI1 V IKXTI)N - !'HARE IN': FRNA"',' NA P

Note:4See Exhibit 21-

T Ti

10 11

1601

744

II

FIGU,1RE 4

DELAY VERSUS CAPACITY -0 HARE TA"7K OFSTT

40-

14 Note:All data based upon

35-*Current (December 1975) ATC*September Baseline Schedule S1*Constant Demand

30- The numbers denote configurations.

A VFR Daily Average

25- 45 FR Daily Average

0 VFR Peak Hour Average

4 0 IFR Peak Hour Average

E Daily 8AM-8PM

~20 Quota Hours 3PM-8PM

IC

14

S aol. TREE ARRIVAL

% E.. RUNWAYS* 04 (see Section 4. 1 1)

6,iAJL, VFR

*101014 5 6 1

CAPACITY(ST)

DELAY VS CAPACITY (ST)DILY AND PEAK HOUR 38

the capacity of the ruriway rcomiirat in. A< i.n e'amp'e,meteorology can affect dclays in ev-i t( ,rist efficientconfiguration at O'Hare. Visual aopr.caches 'n, which the pilotvisually determines his own separatinn from the precodingaircraft) may not be conducted when the cpiling and visibilitylimits fal! below 3,500 feet and n ailes, respect 've y. This,in effect, causes ar iircrc;ased i.pac inq be'tween .- rivals therebydecreasi-.4 rapacity -ind increasi'c !ei .n a As -,ne ceiling andvisibility approach IFR limits (1000/3,, the spacing oetweenarrivals again increases to allow a nreater saifety hufferbetween operations. In x.ndition- f v,=;y ie visiliity, i.e.,less than 500/1, visual observation of 'ho r'n,'ay sys:sem 4S notpossible, requiring additional contrn7li- ca~tion and increaseddependence on pi!ot/controJle- commun iatior. a' I o, whichfurther reduce the efficiency of the -airfe'd system.

The condition of the runways themevc can <-tease spacinq(therefore increase delays by red.0 irj air,_rperformance thus iocreasing runway ')(,(-* panc, t ,oe. T, addit r.snow or ice on the runways will re.o,ire period>i- runway closuresfor maintenance r:: ensure safe npPortina conditKons.

Short term ,henooIi such, a, ground fo-c -, 'he pc-ss.,gc )f afrontal system accomp, n ied oy severe crulenc- can result inthe holding of departures on toe groard -.nd 4ihouno aircraft inholding stacks. These conditions, C'thc hh generally of shortduration, often cause delays of mcjo- oreportiors due o thebacklog of demand created.

E. OPERATIONAL PROC-DlPiES

The term "operational procedures" i-; ;eant to ,.scribe anyprocedural decision -by airport manaqnement or FAA wnicninfluences:

o The selection of on-ratmnal confiq,'raz.icns;o The changina of conf9ur<ionc throughout the

operating day; oro The availability cf rt:owayq for nreratIona, !se.

Several considerations enter into the selection of operational

configurations, most notably meteoloogy and, at. some airports,noise abatement. While wind direct on and velocity ar- keydeterminants in the selection and changing of runwayconfigurations, selection decisions rerain the responsibility ofFAA air traffic control management (multiple -onfgurations canbe used for given wind conditions).

The changing of configurations throughout the operational daycan contribute to relay. Some configuration chirges are used to

adjust runway capacity to demand and create no delay problems.Other air traffic control towers operate under a policy whichrequires periodic runway rotation for noise abatement purposeswhen wind and weather permits. Additional configuration shiftsare required when wind velocity and ceiling and visibilitynecessitate a change in arrival and/or departure runwayorientation.

The unavailability of runways for use due to scheduledmaintenance, construction, and weather related problems, such assnow removal, also contributes to delay. To a large extent,unavoidable weather related problems are the primary reason forunscheduled "down" runways. However, scheduled maintenance andconstruction are a necessary and on-going function of anyairport operation which can contribute to delays. Airportmanagement procedures have not always provided for detailedoperational analyses prior to maintenance and constructionscheduling, and coordination among aircraft and airportoperators and the air traffic control management has not alwaysoccurred to the extent that the delay consequences ofconstruction activities have been minimized.

F. AIRCRAFT OPERATING DEMAND

Simplistically, all airports can be divided into two broadgeneric classifications:

o Origin/Destination; ando Connecting

Origin/destination airports are characterized by largepercentages of passenger traffic either starting or ending theirtrip at the city served by the airport. Some of the passengersmay be making connections and there may be connecting trafficbetween commuter airlines and larger air carriers, but thisrepresents a relatively small proportion of the total traffic.

Connecting airports, on the other hand, are typified by arelatively large percentage of passengers transferring betweenaircraft. This connecting complex role manifests itself in ademand pattern which tends to bunch arrivals and departures inblocks providing the capability to interchange connectingpassengers in a high level of activity during the hours of theday which provide access to the various markets served withreasonable arrival and departure times. The existence of aconnecting complex, with its attendant delay problems, providesbenefits to the extent that:

o An otherwise uneconomical level of service to manycommunities, large and small, is provided by means ofthrough planes and connecting schedules.

I 2 4

o The total level of operations is less trian wojlc otherwisebe required to carry passengers and cargo hetween many citypair markets.

An extreme example of a connection operation is Atlanta'sHartsfield International Airport, os s'own ;n Figure 5. On anhourly basis, operations alternate betw._en predominantlyarrivals or departures. For i given leve' of demand, and agiven runway configurat in, tne relatie ;ix cf arrivals anddepartures will have an effect on the ltav of delay encountered.

IV. ESTIMATES OF AIR CARRIER DELA'S

There is no reporting sys,,-.r,; o, combination of re-crting systemswhich provide information a,)qut sii, ttal dei.ivs in the system andthe possible causes of delay. Da" i fCom Ii. jf t.e systems arediscussed, along with some o f the rrpl icat ons of tht data provided.

A. NASCOM DELAYS

As discussed previously, there are a :.ur-nbr )- diffe-ent systemsfor reporting air carrier delays. The most common'y used systemwithin the FAA is the NASCrOM sstem of reporting air trafficdelays of 30 minutes or U'* e. This system does not provide ameasure of the amount of delay, e.g., total rinutes, but ratheris a measure of the number of operations experiencing aelays of30 minutes or more. To the extent that it does provideinformation, that information is limited to the most severetypes of delays. Total NASCOM delays have been increasing.Between 1976 and 1979, total operations delayed 30 minutes ormore increased from 36,000 to about 62,000 Delays alsoincreased relative to the number of operations, from 3.4 to 5.2delays per 1,000 operations.

While total NASCOM reported delays have been increasing, therehave been no dramatic changes in the causes of NASCOM delays.Data for the four years are summarized in Table 1. Thepredominant cause is adverse weather, accounting forthree-fourths or more of the delays each year.

FTGUT F 5

ATL ANT AGA.A TL

SCHEDULED OPER~ATIONSFRIDAY - AUGUIST 4, i978

ARRIVALS LOCAL DEPARTUREST IME

E00

020304

EZ 05D06E07a

K ZZII z /l IZ08 2

0

1021314131415T

181920

22

TS 65 52 39 26 13 0 0 13 26 39 52 65 78ARRIVALS DEPARTURES

26

TABLE I

NASCOM DELAYS

1976 1977 1978 1979

Total Delays 36,196 39,063 52,239 61,598

Weather 76% 83% 79% 84%

EquipmentFailures 4% ?% 7% 3%

Weather &EquipmentFailures 11% 5% 3% 4%

Other Causes 9% 10% 11% 9%

Total Delays Per1,000 Operations 3.4 3.5 4.6 5.2

While weather may be the primary cause, it cannot be completely

separated from the other factors discussed in Section Ill.

B. STANDARD AIR CARRIER DELAY REPORTING SYSTEM

The Standard Air Carrier Delay Reporting System provides

detailed information about delays classified by type of delay,

but not by cause. The types of delay measured, and their

definitions, are:

27

0 Taxi Out Delays--Determin;;d by mea . ,ring the differencebetween actual taxi out time for an aircraft and apreselected standard fo- each aircraft type and airportoperated out of. The standards developed were based on thefirst ten percentile time of taxi out distributionsconsidering one complete year's worth of operations foreach aircraft type at each airport operated. Whereexperience for partic~ular 3ir'raft types at airports didnot exist, a time relationship was developed andextrapolated from those airports where multiple equipmenttypes were operated,. In no cdse was a standard (as aminimum) to be less than three minutes.

o Taxi in Delays- -Detrmirn! the same way as taxi out de'ays,except that the ;ninimum standard feor any equipment type atany airport would not h- less than Lwo minutes.

o Airborne LCelays--Co;nput,.1 as the difference between actualairborne _Tm-- -off-to-on 1ime) and each respective aircarrier's computer flight plan airborne time - when itexists. The computer flicht plan time considers winds andtemperatures aloft (thus nuillifying their variability), hasan allowance for vectoring in the terminal areas, and is,by policy of each air line, the routeialtitude to oe flown.

Some routes of low stage length do not have a computerflight plan. In these cases, a standard airborne time wasdeveloped based on a linear regression relationship ofairborne time dependent upon route miles as determined fromactual, uncongestd airbornie experience by equipment type.

0 Gate Delay Measurements--Are ,r:ve from each carrier'sdelay codereporting--system, wherein delay times at thegate and delay code, (signifying the reason) are input byairport personnel. In the Standard Air Carrier DelayReporting System qate delav are reported for (1) ATCclearance, (2) weather, (3) ramp congestion, and (4) flowcontrol.

The air carrier data indicate that the average delay per flightincreased from 11.13 minutes to 12.46 minutes from 1977 to1979. Data for each of the four flight phases are summarized inTable 2 for 1977 through 1979.

TABLE 2

DELAY BY PHASE OF OPERATION

Average Delay per Flight, MinutesPhase of Operation

1977 1978 1979

Gate Hold .12 .12 .13

Taxi Out 4.51 4.78 5.07

Airborne 4.27 4.36 4.75

Taxi In 2.?3 2.41 2.51

Total 11.13 11.67 12.46

Since the average delay per flight is calculated byconsidering all flights, including those which experienceno delay, it is worthwhile to investigate the distributionof delays among all flights. Such a distribution isavailable for each phase of a flight from the Air Carrierdata, but not for entire flights. The distribution ofdelays for each of the four flight phases is summarized inTable 3.

z (4

TABLE 3

FREQUENCY DISTRIBITION OF GATE ;IU'D DELAYS,TAXI OUT DEILAYS, AIRBORNE DELAYS, AND TAXI IN DELAYS

EXPERIENCED BY AIR CARRIER FLIGHTS

Percent of FlightsMinutes of G Taxi Taxi

Delay Hold Out Airborne In

0 98.9% 12.2% 39.5% 17.2%

1 0.1 11.9 7.2 26.6

2 0.1 14.3 7.6 22.1

3- 4 0.1 25.6 14.2 21.9

5- 9 0.2 24.7 13.9 9.4

10-14 0.1 7.0 L.q 1.7

15-19 0.1 2.5 2.4 0.6

20-24 0.1 1.0 1.0 0.3

25-29 0.0 0.4 0.5 0.1

30-44 0.1 0.4 0.7 0.1

45-59 0.0 0.1 0.4 0.0

60+ 0.0 (.1 0.4 0.0

100.0 '00.0 100.0

Note: Percentages may not sum to 100.0 due to rounding.

The tables above indicate that we may assume that a clearmajority of flights encounter a total delay below the

average. Combining the results presented in these tables,we note that:

4(1

K

(1) About 99 percent of flights Pncounte- no gate holddelays;

(2) About 64 percent of flights encounter a taxi )ut delayof four minutes or less below the averagc;

(3) About 69 percent of flights encounter an airbornedelay of four minutes or less below the average; and

(4) About 66 percent of flights encounter a taxi in delayof two minutes or less below the average.

The Air Carrier data also enable us to compare the averagedelays at individual airports. The average delays for 1977through 1979 at thirty airports are summarized in Table 4.The 1979 average delay is lower than the 1977 average delayat only three of these airports. The upward treno indelays cannot simply be attributed to higher levels ofoperations, as evidenced by the diversity of changes indelay levels and operations levels summarized in Table 5.

K'

TABLE 4

AVERAGE DELAY PER OPERATION ATSELECTED AIRPORTS, 1977-1979, IN MINUTES

Airport 1977 1973 1979

ATL 10.39 9.93 10.54JFK 9.58 10.75 9.85ORD 9.05 9.33 10.12LGA 8.06 9.17 10.56EWR 7.23 7.70 3.08BOS 7.16 6.72 8.61DEN 6.98 9.23 3.70DCA 6.7? 6.50 6.91PHL 6.45 8.27 6.77STL 5.7q 5.98 6.PcPIT 5.67 5.73 3.84HNL 5.44 5.54 5.75MIA 4.95 5. 5.35LAX 4.87 . 6.45IAD 4.78 4.74 5.54

SFO 4.78 4.50 5.09CLE 4.68 4.48 4.59DTW 4.36 4.59 4.77DFW 4.32 4.61 4.79IAH 4.08 4.83 5.98MSY 3.86 4.39 r.03CHS 3.80 3.67 3.46RAL 3.76 3 4.24IND 3.73 3.91 3.93JAX 3.66 3.62 3.61TPA 3.61 4.17 4.3?SEA 3.48 3.37 4.09RDU 3.47 3.18 3.85MEM 3.20 3.48 4.24MSP 3.14 3.08 3.51

TABLE 5

AVERAGE DELAY PER OPERATION AND TOTAL M RM OFOPERATIONS AT SELECTED AIRPORTS, 1977 AND 1919

(DelTas in Minutes)

1977 1979Airport

Delay Operations Delay Operations

ATL 10.39 508,163 10.54 594,410JFK 9.58 335,473 9.85 325,748ORD 9.05 735,272 10.12 735,524LGA 8.06 351,642 10.56 353,968EWR 7.23 197,404 8.08 211,179BOS 7.16 332,191 8.61 355,959DEN 6.98 452,549 8.70 482,389DCA 6.72 342,001 6.91 351,485POI 6.45 328,746 6.77 344,6307SL 5.79 332,289 6.85 341,854

PIT 5.67 317,779 5.84 346,838 I,HNL 5.44 330,573 5.75 411,376MIA 4.95 313,447 5.35 378,939LAX 4.87 495,312 6.45 547,960IAD 4.78 189,749 .54 176,196SFO 4.78 342,578 5.09 363,791CLE 4.68 241.,28 4.59 265,397OTW 4.36 255,.63 4.77 285,445DFW 4.31 179.903 ,.79 436,362IAH 4.09 ??9, :P3 5.98 295,523MSY 3.86 165, 57 5.03 198,623CHS 3.80 42, t6 1.46 156,427BAL 3.76 74, 96, 4.24 218,66?IND 3.73 :'14,2 0 3.83 212,619JAX V. Q'? 3.6? 132,892TPA 3 61 , 4,3? 241,795SEA 48 4.09 ?08,7?1RDU 3.4' -, 4 f) ,3.R . 206,997MEM 3.20 '2.24 169,069MSP -. 14 t '4P 3.S1 ?8?,750

Finally, the Air Carrier data help present a picture of thetrends in systemwide delays. Various measures of delay aresummarized in Table 6, the main conclusion being thatdelays are generally increasing in number and cost.

TABLE 6

VARIOUS MEASURES OF DELAY AS REPORTED BYSELECTED AIRLINES, 1976-1979

1976 1977 1978 1979

Total Hours of Delay (000's):

Eastern 103 112 119 125

Eastern, United, and American 256 270 296 N/A

Average Minutes of Delay PerOperation:

Eastern 5.9 6.3 6.5 6.7Eastern, United, and American 5.5 5.6 5.8 N/A

Total Delay Cost (000,000's):

Eastern $ 97 $108 S121 $135Eastern, United, and American $221 $258 $225 N/A

Average Delay Cost per Operation:

Eastern $ 93 $102 Sill $121Eastern, United, and American $ 78 $ 89 $ 74 N/A

Gallons of Fuel Consumed inDelay (000,000's):

Eastern 87 92 99 106Eastern, United, and American 228 235 255 N/A

~L

BIBLIOGRAPHY

1. Airline Delay Trends: 1972-1977, FAA-EM-78-11, U.S. Department ofTransportation, July 1978. (Also issues for previous years goingback to 1973).

2. Airport and Airway Congestion: A Serious Threat to Safety and theGrowth of Air Transportation, Aerospace Industries Association ofAmerica, Inc., July 1980.

3. Arad, Bar-Atid, Control Load, Control Capacity and Optimal SectorDsn, Interim Report, National Aviation Facilities Experimental

, Atlantic City, New Jersey, December 1963.

4. Arad, Bar-Atid, Notes on the Measurement of Control Load and SectorDesign in the En Route Environment, FAA, 1964.

5. Attwooll, V. W., "Some Mathematical Aspects of Air Traffic Systems,"Journal of the Institute of Navigation, 30, No. 3, pp. 384-411 (1977).

6. Attwooll, V. W., The Optimum Planning of Air Traffic Flow UnderConstraints, Technical Memorandum MATH 7506, Royal AircraftEstablishment, United Kingdom, September 1975.

7. Bates, G. A., "ASSM-A New Airspace/Airport Simulation Programs,"Journal of ATC, 16 pp. 13-15 (May-June 1974).

8. Ballantoni, J. F., H. M. Condell, I. Englander, L. Fuerres, andJ. Schwenk, The Airport Performance Model, U.S. Department ofTransportation, October 1978.

9. Bobick, J. C. and G. J. Couluris, "Airport and Airspace Delay ModelDescription," FAA-AVP-79-11, U.S. Department of Transportation,October 1979.

10. Braft, R. and S. Mohleji, Analysis of Route Widths in the DomesticAirspace, Report M 63-226, The MITRE Corporation, McLean, Virginia,November 1973.

11. Brauser, K. J., "Measurement of the Control Capacity of ATC Systems,"in Plans and Development for Air Traffic Systems: Papers Presentedat the 20th Symposium of the Guidance and Control Panel; ReportAGARD-CP-18, NATO Advisory Group for Aerospace Research andDevelopment, Brussels, Belgium, May 1975.

12. Condell, H. M. and A. S. Kaprelian, Airline Delay Trends 1973-1974:A Study of Block Time Delays, Ground and Airborne, for Scheduled AirCarriers, Report DOT-TSC-FAA-76-24, U.S. Department ofIransportation, TSC, Cambridge, Massachusetts, 1976.

13. Establishment of New Major Public Airports in the United States, U.S.Department of Transr rtation, August 1977.

14. Hockaday, S. L. M., and A. Kanafani, "A Methodology for AirportCapacity Analysis," Transportation Research, 8, 1974.

15. Horonjeff, Robert, Planning and Design of Airports, Second Edition,McGraw Hill, New York, 19/5.

16. Martin, R. W., "Centralized Flow Control," Journal of ATC, 12,pp. 5-8, September 1970.

17. McNamara, J. E., "The Metering System Concept," Journal of ATC, 18,pp. 24-28, October-December 1976.

18. Medeiros, M. F., Jr., En Route Air Traffic Flow Simulation, ReportDOT-TSC-FAA-71-1, U.S. Department of Transportation, TSC, Cambridge,Massachusetts, January 1971.

19. Ratner, R. S., et. al., "The Air Traffic Controller's Contribution toATC System Capacity," Second Interim Report, Volumes I and II, FAAContract DOT-FA70WA-2142, Stanford Research Institute, Menlo Park,California.

20. Odoni, A. R., and R. W. Simpson, Review and Evaluation of NationalAirspace System Models, Report FAA-EM-79-12, U.S. Department ofTransportation, 1979.

21. Petracek, S. J., and D. K. Schmidt, An Evaluation and Design Criteriafor ATC En Route Sector Configurations, Stanford Research Institute,Menlo Park, California, December 1974.

22. Report of the FAA Task Force on Aircraft Separation Assurance,FAA-EM-78-19, U.S. Department of Transportation, Washington, D.C.,1978.

23. Schmidt, D. K., "On Modeling ATC Workload and Sector Capacity,"Journal of Aircraft, 13, pp. 531-537, July 1976.

24. Siddigee, W., "Air Route Capacity Models," Navigation, 20, No. 4,pp. 296-300, Winter 1973-1974.

25. Siddigee, W., "A Mathematical Model for Predicting the Number ofPotential Conflict Situations at Intersecting Air Routes,"Transportation Science, 7, No. 2, pp. 158-167, May 1973.

26. Siddigee, W., "Computer-Aided Traffic/Airway/VOR(TAC) NetworkMethodologies," Final Report, Tasks 1, 2, and 3, FAA ContractDOT-FA71-WA-2547, Stanford Research Institute, Menlo Park,California, August 1972.

L .... ..... (,

27. Wong, P. J., G. J. Couluris, and D. K. Schmidt, "Aggregate Flow Modelfor Evaluating ATC Planning Strategies," Journal of Aircraft, 14,No. 6, pp. 527-532, June 1977.

28. Air Traffic Control, U.S. Department of Transportation, January 1978.

29. Parameters of Future ATC System Relating to Airport Capacity/Delay,U.S. Department of Transportation, June 1978.

30. New Engineering and Development Initiatives: Policy andTechnological Choices, FAA Contract DOT-FA77WA-4001, Economics andScTence Planning, Inc., March 1979.

31. Harris, Richard M., Models for Runway Capacity Analysis, FAA ContractDOT-FA70WA-2448, MITRE Corporation, December 1972.

32. O'Hare Delay Task Force Study, FAA-AGL-76-1, Volumes I and 11,Federal Aviation Administration, July 1976.

TODAY'S ATC SYSTEM - PROBLEMS AND NEED FOR CHANGE

byRaymond Van Vuren

Director, Air Traffic Service

I have been asked to talk to you today about our conteporary AirTraffic Control System, or ATC System as we call it.

In doing so, I will first trace its historical evolution beforedescribing what it is today. Later, I'll conclude my talk with afrank discussion about some of our concerns about the system.

The evolution of our ATC System began in 1920 when the Post OfficeDepartment established four Aeronautical Stations to support itstrancontinental airmal route. These stations maintained, what can nowbe described, an archaic system of navigational aids, or moreaccurately, beacon lights along the route.

During the late 1920's, some municipal governments began to providesome limited air traffic control at airports by using light signals,or light guns as they are called today.

By the end of this decade, the city of Cleveland had established theuse of radio communications to control airport traffic. Twenty othercities quickly followed Cleveland's example.

Earlier, on May 26, 1926, an Air Commerce Act was enacted. This Act,among other things, directed a civilian agency, the 5opartment ofCommerce, to establish and operate an airway system ard maintain aidsto air navigation. By the end of the year, the first Federal AirRegulations, or FARS as they are called today, were promulgated.

In 1935, methods for controlling airport and airway traffic weredrafted at a conference of government and airway user groups.

A year later, the government assumed operation of the three existingairway traffic control centers which had been operated by the airlinecompanies.

In 1938, the Civil Aeronautics Act added to the responsibilitiesmandated under the Air Commerce Act by establishing the CivilAeronautics Authority, or CAA, which was the predecessor of theFederal Aviation Agency.

The next significant milestone occurred in 1941 when the CAA, for thefirst time, began to operate the airport traffic control towers. Thencame World War II with its tremendous growth in aviation and U.S. airpower. This dictated the need for the common system we have today forhe control of both civilian and military air traffic.

Other milestones were the creation of the Federal Aviation Agency ir1958 and its later consolidation as the Federal AviationAdministration under the Department of Transportation in 1966.

The system of the 30's and 40's, which we term the "First GenerationSystem," relied solely on air/ground radio communications. Itconsisted of an air navigation network and a manual control systembased exclusively on time interval separation between flights.

After World War II, radar was introduced. The addition of thisrevolutionary ATC tool formed the basis for the "Second GenerationSystem"; the system of the 50's and 60's.

During the 1950's, radar was systematiclly deployed throughout thesystem along with other new and improved equipment. Its use providedthe air traffic controller a real-time electronic depiction of theposition of all aircraft under his control.

In general, the more extensive use of radar greatly decreased theamount of airspace assigned to each aircraft, thereby greatlyincreasing the efficiency of the system.

In the early 1960's, the Air Traffic Control Radar Beacon System(ATCRBS) was introduced. This system, which consists of airborneradar beacon transponders and ground-based radar beacon interrogatorsand receivers, formed the basis for today's system, tne "ThirdGeneration System."

In retrospect, you can now see that our ATC system evolved from, andis based on, three unique type facilities, plus a fourth which I willshortly identify.

First, there is the Air Route Traffic Control Center (ARTCC) whichevolved from the Airway Traffic Control Center operated by theairlines.

Second, there is the Airport Traffic Control Tower (ATCT) whichevolved from the early control towers which were operatedindependently by local government.

Third, there is the Flight Service Station (FSS) which evolved fromthe aeronautical stations operated by the Post Office Department.

Lastly, there is a fourth unique type facility, the Air TrafficControl System Command Center, which was established in 1970. Thiswas in respone to isolated clusters of congestion which occasionallydisrupted the system.

I will now, in the brief time that we have, attempt to describetoday's system-a system of facilities.

Today's system, the "Third Generation System," the system of the 70's

and early 80's, is generally organized around the ARTCC, or Center, aswe call it. These facilities provide separation service for allaircraft operating under Instrument Flight Rules (IFR) in airspacegoverned by the U.S.

The airspace has been divided into 25 geographical areas. Each ofthese areas is under the jurisdiction of a center. Twenty residewithin the conterminous U.S. (CONUS); the remaining five in Alaska,Hawaii, Puerto Rico, Guam, and Panama.

The original center boundaries were generally drawn to conform withthe geographical boundaries of the states within a particular FAARegional Office's jurisdiction. This was intended to facilitatedetermining regional airspace configuration, along with other factors,makes needed changes in national airspace difficult to implement.

To provide a manageable system within the Center, the airspace isfurther devided both horizontally and vertically into blocks ofairspace called sectors. The sectors are classified by altitude:ultra high altitude at and above flight level 370; high altitude atand above flight level 180; low altitude below flight level 180.Other airspace, usually below 7,000 feet AGL and within 35 miles of aterminal radar facility, is released to terminal approach controlfacilities.

It will be helpful to you at this time to note that altitude at ancabove 18,000 feet MSL are called Flight Levels. They are based onaircraft altimeter readings set to correct for a constant mean sealevel barometric pressure of 29.92 inches of mercury.

The vertical aircraft separation minima applied by the center is basedon altitude level: A 1,000 feet minima up to and including flightlevel 290; a 2,000 feet minima above flight level 290.

When not applying vertical separation, the following horizontal radarseparation may be applied:

Below flight level 600 - 5 miles

At or above flight level 600 - 10 miles

You should also note at this time that aircraft operating above 12,500feet MSL are, by regulation, required to have an operating Mode Caltitude reporting transponder.

Additionally, a Positive Control Airspace (PCA) has been establi.ned

for approximately all airspace at and above flight level 180 in theCONUS. In addition to the Mode C transponder, the aircraft mustoperate under an IFR flight plan.

The air traffic control positions of operation within the Center are,quite naturally, also called Sectors. They are usually situated suchthat adjacent sectors control adjacent airspace.

The controller positions of operation at each sector are typically the"Radar Controller," "Data Controller," "Assistant Controller," and"Coordinator."

Through the use of full digital radar displays called plan viewdisplays, or PVDs, computer input devices, air/ground radiocommunications, landline communications between other sectors,adjacent centers and terminal approach control facilities, and paperstrips called flight progress strips, the radar controllers separateand provide services sector-by-sector as the aircraft traverse theCenter's airspace.

The PVD is driven by an ATC automated system called NAS En Route StageA. This system, which is an IBM 9020 computer, is in operation at all20 Centers within the CONUS. It is a large and complex system whichaccomplishes many functions simultaneously. These functions consistprimarily of Radar Data Processing (RDP), Flight Data Processing(FDP), Radar Simulation (DYSIM), and Data Recording.

The FDP function performs flight plan composition and storage, routeprocessing, updating, and automatic transfer of flight data within theCenter, between the Center and the adjacent Center, and the Center andits associated terminal facilities.

The RDP function performs primary radar and beacon target processingof target reports received from our Radar Digitizers (CDS) aL theLong-Range (ARSR) sites, precise tracking of aircraft, automatichandoffs of aircraft between the sectors, adjacent centers andassociated automated terminal facilities. It displays these data astarget symbols, history trails, and data blocks. The data block isautomatically associated with the aircraft target and containsessential information, such as aircraft ID, assigned altitude,reported altitude, and ground speed.

To assist the controller in maintaining an equitable distribution ofdelays and an optimum arrival interval between landing aircraft at ourmajor airports, en route metering is employed. This function, whichis now accomplished manually through controller judgements, will soonbe automated. Precise calculations of the time and speed at which anaircraft should depart an outer metering fix will be made based onreal-time flight data. Arrival sector controllers will receive thisinformation in the form of sequenced aircraft arrival lists andmetering fix crossing times displayed on their PVDs.

The controller wil, in turn, adjust airspeed or radar vector to ensurethat the aircraft crosses the metering fix at the optimum time. The

computer program will be delivered to the field in December. After a90 day test, evdluation and controller training period, we expect thefirst center to be operational in March, the last in July.

There are a multitude of other en route functions such as flowcontrol, en route metering, search and rescue coordination, militarymission coordination, training, and the like that would require a muchlengthier discussion.

Also, Military Training Routes (MTRS), Military Operating Areas(MOAS), Perferential Routes (PDARS, STARS), and profile descent aboutwhich time does not permit me to discuss.

As you recall, I mentioned that there were also five Centers outsidethe CONUS. Three of these low-activity, offshore Centers haverecently received the EARTS system.

This system is essentially an ARTS-IlIA, which I will shortlydescribe, but has been modified to accept digitized radar data fromany combination of up to five long-range (ARSR) or short-range (ASR)radars. Another notable exception is that it employs full digitalradar displays, or PVDS, in common with the other larger centers.

The next facility in our system of facilities is the Airport TrafficControl Tower, or tower, as we call it.

The tower provides ATC service in the vicinity of airport. Wheretraffic volume is moderate to heavy, the facility consists of both thetower and a Terminal Radar Approach Control called a TRACON.

Today, there are approximately 500 towers of which about 50 percentare associated with either an integral or separate TRACON. Others areprovided radar approach control service by the center sectors whereradar coverage permits.

Today's towers fall into three categories: the VFR Tower, the NonradarApproach Control Tower, and the Radar Approach Control Tower.

The VFR Tower essentially consists of three positions of operation, orcontrol positions, as they are called at terminals, they are "FlightData", "Ground Control," and "Local Control." The latter positionprovides visual control of aircraft while they are departing, arrivingor flying in the local traffic pattern; the former areself-explanatory.

The functions of the VFR tower are sequencing arrival and departuretraffic and the separation of aircraft on the ground.

The nonradar approach control tower differs from the VFR tower in thatit provides separation services to aircraft operating IFR through

precise ATC procedures. It accomplishes this through the addition ofan approach control position. The approach controller separatesaircraft by assigning various nonconflicting routes, altitudes, ahdtime restrictions to aircraft under his or her control.

Lastly, there is the radar approach control tower. These facilitiesare established in terminal areas that have moderate to high trafficdensities where radar can expedite the flow of traffic.

This facility consists of a tower cab, which is essentially the sameas the VFR tower, with the addition of a Brite radar display, and,most importantly, a TRACCN.

The TRACON consists of e large room within which the radar operationis housed and conducted. A number of radar approach control towersare what we term TRACABS. In the TRACAB, the approach controlfunction is performed using a Brite radar display in the tower ratherthan in a separate room.

The TRACON is, in essence, a small center which provides ATC and otherradar services in the terminal airspace delegated to it by theCentaur. This airspace is usually that airspace within 35 miles cf theprimary airport and below 7,000 feet AGL. The exceptions to this ruleare the larger TRACONS such as the New York aid O'Hare TRACONS. Eachof these larger TRACONS control traffic into and out of a number ofprimary and satellite airports.

Its typical positions of operation are "Approach Control," "DepartureControl," "Flight Data," and "Coordinator," The busier may include,in addition to mulilple positions, "Parallel Approach Monitors."These latter positi(ns monitor aircraft conducting simultaneousinstrument approaches to parallel runways.

The TRACON controllers provide essentially the same type of service astheir counterparts at the Center. The significant difference beingthat the service is provided within smaller divisions of airspaceusing reduced aircraft separation minima and a high degree ofvectoring.

The following separation minima is applied in the terminal environment:

Horizontal Radar Separation

5 miles if aircraft 40 miles or more from the antenna

3 miles if less than 40 miles from the antenna

1 1/2 miles for participating VFR aircraft when within 15

miles of the antenna

Vertical Separation

IFR Aircraft - 1,000 feet

Participating VFR Aircraft - 500 feet

A particiating VFR aircrft is one that is voluntarily receiving radarseparation services within a Terminal Radar Service Area, or TRSA, aswe call it.

If the VFR aircraft is operating in a Terminal Control Area (TCA) theservice is mandatory. The TCA is, in effect, an area of positivecontrol airspace surrounding a major airport.

We have, to date, established 22 TCAs and 135 TRSAs.

All of our terminal radar facilities include some degree of automationdependent on traffic loads; ARTS-Ill and ARTS-IlIA at our high trafficdensity facilities, ARTS-II at our moderate traffic densityfacilities, and TPX-42 at our light traffic density facilities.

It may be helpful to point out at this time that air carrier passengerloads are appropriately factored in when determining which facilitiesget which equipment.

ARTS III, which provides a fairly high degree of automation, is inoperation at 64 of our major terminal facilities. It consists of aData Acquisition Subsystem (DAS), Data Processing Subsystem (DPS), anda Data Entry and Display Subsystem (DEDS).

Simply put, this system provides the terminal controllers the samekinds of information that NAS en route Stage A system provides theiren route counterparts. However, there are some notable exceptions.

The information, that is, the target symbols and data blocks, arepresented to the controller on time-shared radar displays, or, PlanPosition Indicators (PPIS), along with br- lband primary radar andbeacon targets. Consequently, two systems, the old system and the newsystem, are concurrently in operation. This provides an immediatebackup in the event of an ARTS System failure.

Another exception is that the basic ARTS III System does not receiveprimary radar information from nontransponder equipped aircraft.Consequently, these aircraft are handled in the "Second Generation"manner using the concurrent broadband display feature.

ARTS IlIA is an enhancement of the basic ARTS III System. It includesprimary radar target tracking, autorecovery after a computer failure,and Continuous Data Recording (CDR).

'44

This enhancement will be installed at 26 of our major terminalfacilities. It is now in operation at the Tampa TRACON and will soonbe commissioned at the second planned site, the New York TRACON.

ARTS II, which provides a somewhat lesser level of automation, is inoperation at approximately 70 of our facilities with more on the way.

Again, it can best be described by comparing it with a previouslydescribed system, the basic ARTS III. It is essentially the sameexcept that it performs no tracking. Target symbols and accompanyingdata blocks are associated with the Beacon targets at the reportedposition (Radar X-Y) on a scan-by-scan basis. If the target report ismissing; e.g., a garbled Beacon reply, the alphanumeric data is frozenat the last reported position rather than displayed at the predictedposition as in ARTS III.

Lastly, there is the TPX-42 System. This sytem is in operation at 35of our lower activity terminal facilities and also a large number ofmilitary ATC facilities.

The TPX-42, in essence, a smart beacon decoder system which includes avery low level of automation. It receives and decodes beacon repliesand displays the resulting data as target symbols and limited datablocks. The data block consists of the numeric octal beacon code andreported altitude if a Mode C reply was received.

These systems will be eventually replaced by ARTS I.

In addition to the features that I have very generally described, mostof our automation systems include some recent safety enhancements:

Conflict alert in NAS en route Stage A and ART 11, Minimum SafeAltitude Warning (MSAW) in ARTS III, and Low-Altitude Alert System(LAAS) in TPX-42.

These safety enhancements are also planned for our ARTS II System infuture budgets.

Unfortunately, time does not permit discussing further functions andaspects of our terminal ATC System.

The next of our unique type facilities is the Flight Service Sttion,or FSS. These vital facilities have the prime responsibility forconducting preflight breifings, en route communications with VFRflights, assisting lost VFR aircraft, origniating Notices To Airmen(NOTAMS), disseminating aviaion weather information, accepting andclosing flight plans, monitoring radio navigational aids (NAVAIDS),participating with search and rescue units, and last, but not least,operating the national weather teletype systems.

L,

In disseminating en route weather information, the FSS employs anEnroute Flight Advisory Service, or EFAS, as we term it. This servicewas first implemented on the West Coast in 1972 to provide real timeweather reports directly to pilots inflight on dedicated frequencies.This service, which has reduced weather related accidents, is nowprovided at 44 locations.

EFAS facilities will be equipped with digitized weather radar displaysin the very near future.

We are now embarked upon a major FSS automation modernization programto improve service to the users.

As an essential element of this program, we have reviewed the numberand location of existing FSSS with a focus towards consolidation.Upon completing our review, we have tentatively selected 61 FSSS toreceive automation. Factors such as general aviation activity,airport operations, and based aircraft were considered in ourdecision. These locations are centers of general aviation activityand not necessarily the same location as the existing FSSS.

The automated FSS, which will be housed in newly constructedbuildings, will be phased in over the next 5 to 7 years. It willintroduce a number of new capabilities which will utltimately providespecialists with rapid access to weather information and '-he abilityto provide instant processing of flight plans. Additionally,telephone "bottlenecks" will be eliminated by enabling the user toaccess the system directly to obtain briefings and file flight planswithout the assistance of a specialist.

There are currently 319 FSSS in operation. They range from very small"One-Man" facilities to large facilities employing more than 100people.

Many positions of operation are involved in the operation of the FSSsystem which would require a separate briefing.

Lastly, there is the Air Traffic Control System Command Center(ATCSCC). This single facility is located in our WashingtonHeadquarters building.

It consists of five distinct operational components: Central FlowControl, Airport Reservations Office, Central Altitude ReservationOffice, Contingency Command Post, and a National Weather DisseminationService.

Central Flow Control is accomplished by a Central Flow ControlFacility, or CF2, as we call it. This facility is intended to, ineffect, manage the flow of air traffic throughout the nation to ensureoptimum efficency, safety, and economy.

To enable its operation, i! Pmproyan a-,'a )f eq,,ipment andcommunications outlets, in lujirg i remote computer complex atJacksonville, Florida. Untortunately, this eq;jipmrent has yet to bedeveloped to the extent necessary f-r the Central Flow Function to beperformed as intended.

The Airport Reservation Office (APO) was established to relievecongestion at four of our busiest airports: O'Hare, Kennedy, LaGuardia, and Washington National. It accomplishes tnis task byrequiring reservations during high traffic activity hours to reducethe activity to a more efficient level.

The Central Altitude Reservation Office (CARF) exists primarily tosupport the military and assist Local Flow Control specialists at theCenters. It accomplishes this by providing detailed, easilyinterpreted assessments of pending military exercises, including airrefueling operations, missile launchings, and the like which requirespecial airspace protection.

Additionally, CARF serves as our coordinating agency for processingairspace reservation requests which will traverse foreign airspace.

The Contingency Command Post (CCP) becomes activated when a majorcalamity occurs somewhere within the ATC system.

Examples of a major calamity would be a catastrophic failure at one ofour Centers or a widespread, and I might add, illegal controllers'strike.

The last of the functions of the Command Center is its weatherservice. Specifically, its assembly, analysis, and dissemination ofexisting nationwide weather reports. This includes upper airsituation, orientation of the jet stream, areas of clear airturbulence, and forecasts of terminal weather conditions.

In concluding this part of my talk, I would like to throw some numbersat you that should assist you in appreciating the size of our AirTraffic Control System and the extent of its services.

Air Traffic Control Specialists 21,000Flight Service Specialists 5,000Expert Technicians 10,000ARTCCs 25ATCTs 431 (FAA)FSSs 319Air Route Surveillance Radars 124Airport Surveillance Radars 195Radio Communications Air/Ground 562VOR/TACAN/DME 1,000Instrument Landing Systems (ILS) 701Direction Finders (DFs) 203And More Coming

I will now conclude my talk with a frank discussion of some of ourmore pressing concerns.

The first of these concerns is airspace utilization. Capacity andefficiency remain a challenge in our management of navigableairspace. Although better use is being made of this limited resourcethan ever before, we have yet to take advantage of available option,.

An excellent example of this is our inability to fully implementexisting concepts, such as Area Navigation (RNAV) and great circleroutes. As a result, we have been subject to criticism from users whohave made substantial investments in avionics which provide thesecapabilities.

A second concern is that airspace saturation in major terminal areaswill become more severe unless runway acceptance rates can be safelyincreased. Many possibilities have been examined, including theconstruction of new airports. This option does not appear viable atmost hubs in the foreseeable future because of the high cost ofenvironmental concerns.

These concerns underscore the need to expand the capacity of theexisting airports. New metering and spacing techniques must bedeveloped along with techniques to cope with the problem of wakevortices. The highest priority should be given to developing theneeded technology to increase runway acceptance rates and, thereby,airport capacity.

In respone to our concern about the overall efficiency of the NationalAirspace System, we are proposing a complete airspace review to beginin the very near future. This review is prompted by the changedaviation demands since our current airspace structure was firstestablished. The following factors were considered in determining theneed for this action:

First, the current airspace system has never beensystematically reviewed.

Second, there is the recently expressed user and publicanxiety over the efficiency and safety of the system.

Third, to reduce our over dependence on foreign energyresources, we must develop a less energy intensive system.

Fourth, the intended use of the airspace by the various usergroups has shifted considerably in the recent years. Thegeneral aviation aircraft is more numerous and sophisticated,air carrier and commuter operations have greatly increased,and deregulation has increased airspace loading in the lowaltitude stratum.

Lastly, we must better acconnodate mi~itary demands forspecial use airspace and high-speed low altitude trainingroutes.

To accomplish these oh~ectives, we propose the foliowing three phaseaction plan:

Phase I - A 30 month evaluation of the airspace structureat FL 180 and above.

Phase II - A 48 month evaluation of the airspace structurebelow FL 189.

Phase III - A final 120 month period during which thefollowing new, programs, systems, and procedures will beintegrated into the National Airspace System.

Discrete Address Beacon System (DABS)/Dita LinkAutomatic Traffic Advisory and Resolution Service (ATARS)Metering and SpacingAutomated FlowConsoliiatien of ATC Facilities9020 Replacement and EnhancementsReduced Separation MinimaArea Navigation (RNAV)Other R&D Initiatives

Other concerns are in the new facilities and equipment area (F&EProgram). Management of our flight service station modernizationprogram is one of our most frustrating challenges. Systemwide, theoperation of the FSS is inefficient because we are obligated tooperate small facilities instead of using our resources in the largerfacilities where demand car more efficiently be met. Our delays inestablishing a consisten course for improvement and false starts mayhave diminished the dedication of our FSS people and adverselyaffected the users of their service.

Our major automation systems, NAS en route Stage A and ARTS 11, havefulfilled all expectations, but near-term modifications andenhancements are needed to keep pace with air traffic growth andincreasing user demands until these systems are replaced. Although wehave experienced several isolated computer capacity problems at acouple of our sites, they have been effectively resolved through minorsoftware design changes, contrary to self-serving charges by thecontrollers' union, the number of computer outages in the system hasbeen continually on the decline. Moreove, we consider the presentrate of computer outages reasonable in both number and duration andour backup systems fully adequate.

Another of our major concerns and one that involves safety is the

continuing acceptable rate of errors in the system. These errorsinvariable are the result of improper coordination, lack ofcoordination, inattention, forgetfulness, or inadequate positionrelief briefings.

During the five-month period from January to May of this yer, therewere a total of 247 system errors reported. Although this representsa 7 percent decrease over the same reporting period of last year, itis, nonetheless, unacceptable.

While we have developed and implemented training programs to correctthese errors, it is apparent that the human factors element needs morpattention in planning system improvements. Towards that end, we areworking with the MITRE Corporation to develope programs desionec toaddress this weakness.

Even with better planning, quality control, and management oftechnology, safety in the system will, in the near-term, cotinc t)rest heavily on the shoulders of the controller and ultimately n(lpen'!upon the motivation, dedication and competence of our supervisory andcontroller workforce.

Despite the foregoing, we must be prepared for the possibility o- anillegal controllers' strike. To ensure the continuation of ATCservice should it occur, we have developed an operational contingencyplan. It is an action plan to provide the best service possible usingour remaining personnel. While the plan is in effect, our capabilitywill be severly reduced since our facilities will be manned primarilywith supervisory personnel. Consequently, ATC operations will also beseverely reduced to ensure safety.

The plan provides for transportation of the most people possible and,therefore, preference is given to air carrier and air taxi flights.Long-range flights will be given top priority as alternate travelmethods are least acceptable and available for long distances.

Permanent daily schedules will be provided to limit traffic to thecapacities that will be available at the facilities servicing theseflights. Routes from point to point are predetermined and specificaltitudes are assigned to each airway. This was done to provide, tothe extent possible, built-in altitude separation, a smooth flow oftraffic in all the facilities, and no airborne delays.

The schedules contained in the plan will be based on current airlineflight schedules; however, departure time will be altered. All but aminimum number of long-range flights that traverse more than 500 mileswill be accommodated. International flight departure and arrivaltimes have been adjusted to the minimum extent possible and all areexpected to be accommodated except some serving Canada and Mexico.This approach will ease the air traveler's burden and airport waitingarea congestion.

As a result of this plan, airlines will be required to accept routesand altitudes which are far less than desirable. In addition, thepredetermined schedules must be followed.

Some other air carrier and air taxi flights will be accommodated butwill be restricted to the extent necessary to avoid impacting thelong-range flights.

To provide for these basic transportation needs while the plan is ineffect, the military will be asked to terminate some training andother noncritical flights. Military necessity or emergency activitieswill continue to receive top priority and be supported to the extentpossible with our limited resources.

All other types of air traffic will be impacted to the extent requiredto maintain essential services. We anticipate the need to establishspecial restrictions on general aviation activities and curtail orsuspend services at specified tower controlled airports.

The operational contingency plan will also address several otherrelated problems such as probable sabotage, threats or physical harmto nonstriking controllers and supervisors, and special emergencyrestrictions. These restrictions include disapproval of most VFRflights in terminal control areas, refusal to accept some proposed IFRflight plans and termination of TRSA service.

We will keep the air traffic control system operating safely, servingthe critical requirements of the military, emergency flights and mostnecessary long distance air carrier flights. All other users of thesystem will be severely impacted.

In conclusion, and on the brighter side, we are confident that ourcontroller intensive system, a factor which is at the heart of most ofour system problems, will be increasingly less so intensive in thelong-term through automation enhancements. Significant among theseare Discrete Address Beacon System (DABS), Automatic Traffic Advisoryand Resolution Service (ATARS), and Metering and Spacing.

The introduction of DABS into our ATC System, coupled with its two-waydata link and improved ability to discretely identify aircraft, willprovide the capability for the introduction of ATARS. ATARS willgreately reduce the midair collision possibility through its cockpitdisplay or proximate aircraft and conflict resolution commands.Eventually, we expect that with the advent of metering and spacingcoupled with the data link, we will, in effect, automate today'smanual controller vectoring procedure and establish safe and optimumflows of air traffic throughout much of the system.

As these systems are proven and increasingly in use, the controllersrole will gradually shift to that of a system manager/monitor ratherthan the essential element in the system.

Thank you for providing me an opportunity to talk with you today.

I will now conclude my talk with a frank discussion of some of ourmore pressing concerns.

The first of these concerns is airspace utilization. Capacity andefficiency remain a challenge in our management of navigableairspace. Although better use is being made of this limited resourcethan ever before, we nave yet to take advantage of available options.

An excellent example of this is our inability to fully implementexisting concepts, such as Area Navigation (RNAV) and great circleroutes. As a result, we have been subject to criticism from users whohave made substantial investments in avionics which provide thesecapabilities.

A second concern is that airspace saturation in major terminal areaswill become more severe unless runway acceptance rates can be safelyincreased. Many possibilities have been examined, including theconstruction of new airports. This option does not appear viable atmost hubs in the foreseeable luture because of the high cost ofenvironmental concerns.

These concerns underscore the need to expand the capacity of theexisting airports. New metering and spacing techniques must bedeveloped along with techniques to cope with the problem of wakevortices. The highest priority should he given to developing theneeded technology to increase runway acceptance rates and, thereby,airport capacity.

In respone to our concern about the overall efficiency of the NationalAirspace System, we are proposing a complete airspace review to beginin the very near future. This review is prompted by the changecaviation demands since ou- currant airspace structure was firstestablished. The following factors were considered in determining theneed for this action:

First, the current airspace system has never beensystematically reviewed.

Second, there is tne recently expressed user and publicanxiety over the efficiency and safety of the system.

Third, to reduce our over dependence on foreign energyresources, we must develop a less energy intensive system.

Fourth, the intended use of the airspace by the various user

groups has shifted considerably in the recent year Thegeneral aviation aircraft is more rumerous and sophisticated,air carrier and commuter operations have greatly increased,and deregulation has increased airspace loading in the lowaltitude stratum.

AN OVERVIEWof the

FAA ENGINEERING ANW-VELOPMENT PROGRAM

byRobert W. Wedan

Director, Systems Research and Development Service

The Federal Aviation Administration Engineering and Development program isdriven by the need to maintain and enhance the safety of the aviation system,to achieve improved performance of the system for the participants and theflying public, and to make the system more productive--to constrain the costof the system to the Nation.

The primary driving forces in the SAFETY area are:

o To improve the FAA air traffic control separation and advisory service andreduce the risk of midair collisions (with a minimum economic andoperational burden on the users).

o To achieve improvements in the air traffic control system and in aircraftoperations which take into full account the limits and capabilities of themen and women operating the system.

o To reduce the number of accidents which occur during the approach and

landing phase of aircraft operations.

In the area of PERFORMANCE (Capacity/Delay), the goals are:

o To achieve a more energy efficient system through reduction of ATC delaysand increases in airspace and airport capacity, especially at and nearmajor hub airports where expected increases in operations and enplanementsmust be accommodated.

o To reduce the impact of delays due to weather.

o To improve airport pavement by application of new technology to itsdesign, construction and maintenance.

o To gain a better understanding of factors such as demand/capacity/delayrelationships and the impact of specific improvements in terms of safetyeffectiveness, real costs, and other benefits.

o To examine longer range technological alternatives for system improvementsand innovation so as to permit evolution to new capabilities in a timelyfashion.

o To reduce negative impacts of aviation on the environment, and to

accommodate the impact of aviation activity on landside congestion.

p

For PRODUCTIVITY, the needs are:

o To reduce the labor intensiveness of the FAA air traffic control systemand the facilities maintenance system by the better use of people and theuse of automation.

o To achieve improvements in FAA's communications, navigation, andsurveillance systems and to explore the possibilities of systemsintegration to achieve a better system at lower costs to both theGovernment and the users.

,Vt!-GRAPH -- PRESENT PROGRAM FOUNDATIONS)

What are the sources of requirements and objectives for the FAA Engineerinqand Development progrdm? ir summary, they are:

o Once aqiin, the first ),, FAA's overall aviation safety priorities.

0 The sec _nJ aro s;the- ,yc,tom performance measures: near-midair collisions,system errors, ot error ., Hlays, etc.

o We .!raw hjavr ',, - -uo. nn the views and requirements as expressed byFAA s r r ,,ystein , .. i,- r, - g(,jjlators.

o , , , 'e ,-' community continually. One way of doingt .. .j. N ,: W1Fiuieering and Development Initiatives processO, , , , t', 1.7 -79 time frame, and a large series of

4i 'a:, , . v , *e vmmendations was derived from that process.

1 F) i .; •,d, ) the t!,rn of the century to try to foresee'v' ,,.,: * .r- ' tor 'n that time frame, and to tailor our EngineeringInr VA- Tnf , i, , ndiy to meet the requirements of the Year 2000ind beynJ W r1,1 1, e many issues yet to be resolved, we believethat i deve'opmon, - I ihn,i p i s necessary to assure that we and the,'Ommuf"y -st , tr, fea- ?000 needs and move to meet those needs.

,,-GRAPH -- AVIATION FORECAST CHART)

What is trie gr'r.wttn lackq-;und igainst which we are working? It can becharacterized in many ways, but it may be easiest to say that we expectcontinuPd moderate jowth in the system over the next 20 years, with quitedramatic changes in the aircraft mix. Even with moderate growth ofconventional scheduled ,ir carrier operations, demand is likely to outstripcurrent capacity. Commuter air carrier operations have grown at the rate of20 percent a year for the last two years, and we can expect general andbusiness aviation to grow rapidly, with resulting changes in the nature of thesystem.

FAA programs fall into several categories:

(VU-GRAPH -- CURRENT DEVELOPMENT ACTIVITIES)

o There is a series of development efforts primarily aimed at near-termsystem improvement. The list of current development activities will giv

you a feeling of specific programs of relatively near-term work in the

Engineering and Development organization.

(VU-GRAPH -- ADVANCED DEVELOPMENT ACTIVITIES)

o There are activities in advanced development. Again, there is a group ofspecific programs.

(VU-GRAPH -- BRIDGING ACTIVITIES)

o There is a third category of "bridging" efforts that impact current majorsystems development, advanced development, and system integration.

Let me now try to group this set of programs and efforts into more digestibleform. (Each underlined item in what follows represents a specific programarea.)

o A major FAA effort centers around Aircraft Separation Assurance. Withinthis program are the completed Conflict Alert program and the developmentof a conflict resolution system.

The Discrete Address Beacon System (DABS), which will provide a majorimprovement in FAA's air traffic control surveillance capability andwill make possible automatic digital communications for safetyservices, is at the heart of the Aircraft Separation Assurat eprogram, as well as a number of other parts of the system evolution

(VU-GRAPH -- DABS/ATARS CONCEPT)

The DABS system will make possible the Automatic Traffic Advisory anoResolution Service (ATARS)--a ground-based collision prevontionsystem.

(VU-GRAPH -- DATA LINK)

The DABS data link will provide a series of services beyond ATARS:clearance confirmation, severe weather warnings, accurate informationon winds aloft for flow management, and others.

(VU-GRAPH -- ACTIVE BCAS)(VU-GRAPH -- FULL BCAS)

The Beacon Collision Avoidance System (BCAS)--an airborne system toprovide last ditch protection against collisions--is in a late stageof development after many years of work. The Active BCAS, and laterthe full-capability BCAS just beginning development, will provideprotetion particularly in areas where DABS and ATARS services arenot available. BCAS may be particularly valuable wherever systemerrors and blunders are a problem, and where air traffic services maybe limited.

The DABS data link will be used to provide coordination and controlfor BCAS in areas where it cannot function optimally.

(VU-GRAPH-- AUTOMATED TERMINAL SERVICE)(VU-GRAPH -- AUTOATED PILOT ADVISORY SYSTEM)

-Another prospective weapon against midair collisions is the use ofautomation at the low-activity airport, the airport which may not yetqualify for VFR tower service. Both FAA and NASA have been workingin this area, on concepts called Automated Terminal Service (ATS) andAutomated Pilot Advisory System (APAS), and a program which willextend these two successful feasibility developments is underconsideration at the moment.

Finally, an effort which grew out of the New Engineering andDevelopment Initiatives activity is called Alternate SeparationConcepts. The New E&D Initiatives people felt it important for FAAto examine new ideas by which especially general aviation couldachieve certain safety benefits for themselves and in relation tofull IFR aircraft, without being full-time participants in the ATCsystem. An effort is underway, working with the community, to try toput some flesh on those still-bare conceptual bones.

(VU-GRAPH -- DEVELOPMENT ROADMAP FORCOMPUTER REPLACEMENT PROGRAM)

0 A second major area is that of the replacement of the air traffic controlcomputer system and the efforts to introduce additional automation intothe ATC system. First, and most immediately important, is the program toreplace the existing ATC computer system, starting with the en routesystem andthFe display suite, with a far more capable, reliable, highercapacity system--one that not only performs current ATC automaticfunctions well, but which will serve as the foundation for a series ofadditional automation steps. The knowledge and wisdom FAA can apply tothe design and architecture of the computer replacement program will, inlarge measure, determine the degree and the pace at which further systemimprovements can be achieved. You will hear more later on about thecomputer and display replacement program.

(VU-GRAPH -- ETABS)

o Other work in advanced automation for improved safety and productivityincludes the completion of an En Route Minimum Safe Altitude WarningSystem (EMSAW), the development of electronic tabular displays for bothen route and-terminal operations, advanced en route automation startingwith the en route meterin 9 system currently in a late stage ofdevelopment, the Automated En Route ATC System concept (AERA), and thedevelopment of an integrated Flow Management system which will tietogether a series or developments and will use automation tools to helpcontrollers provide optimum terminal airspace and airport utilizationwhile permitting aircraft to operate in a fuel-efficient manner.

o The FAA Weather program is aimed at severe weather detection, processing,display and coomunication to pilots. The efforts include a future weathersystem design which will encompass an improved nationwide Doppler weather(NEXRAD) network to be shared by several U.S. agencies. Efforts to

improve short-term forecasting, work to better show weather on terminaldisplays, work to provide automatic weather observations both at majorairports and very small ones, and developments to permit the use of DABSdata link to provide better information to pilots and to draw informationfrom aircraft for flow management, are part of this area.

0 In addition to the weather activities which involve radar services, a taskforce is underway to examine the future requirement for surveillance--bothprimary and secondary radar. While there is little question that primaryradar will be a long-term requirement for terminal operations, there aresome who believe that primary radar should not have a long-term role fortarget acquisition in the en route system, but that secondary radar shouldbe the primary surveillance source. This is a large and old issue whichrelates closely to the question of universal carriage of secondarysurveillance radar transponders. FAA's effort here is to assess thesituation once again and to offer a series of alternatives to top FAAmanagement for consideration.

o One FAA program area deals with airport capacity and delay froblems, noisereduction and fuel conservation. Through the use of workable airportcapacity and delay models now completed, FAA/industry task force-s-Thelped FAA draw bottom lines to a series of development efforts. Theseefforts have helped FAA to pinpoint and concentrate on those improvementareas which can make a significant improvement in airport capacity andreduction of delay. The most important among them are improved trafficflow management through the use of a nationwide Integrated Flow ManagementSystem which will use both ground system and aircraft capabilities tocreate a traffic flow pattern which matches airport capacity to the flowsystem, which can accommodate airborne area navigation and performarcemanagement systems, which can utilize the most efficient safe separationstandards, and which can cope with the problems of wind shear and wakevortex.

(VU-GRAPH -- ASDE)

A significant part of this work is the optimization of the airport itselfin terms of airport layout, exit design, runway occupancy time control,optimum utilization of short runways for commutcrs and general aviationaircraft, the successful deveopment of a modern airport surface detectionradar (ASDE), automated surface traffic control where this can beJustfied, and work to improve airport environment and access forpassengerc.

o FAA's program to characterize and ameliorate the risk of wind shear ha-yielded important results. A network of detectors nas now bteninstalled. While we have learned a great deal about the wake vortexproblems in work conducted by FAA on detection and avoidance of vortices,and NASA has done considerable work on wake vortex alleviation at thesource, we have not yet achieved an operationally acceptable wake vortexdetection and avoidance system, and work continues.

o An effort is underway to dramatically enhance FAA programs relating topilot and controller performance enhancement and error reduction--the areaof human factors. A number of activities are underway, and we are workingto strengthen this effort in important ways. Among the activitiescurrently in work are efforts to establish cockpit informationrequirements for the future ATC system; to define improved cockpitalerting, warning, an ight phase monitoring systems; to evaluafThT1impact of new navigation systems on pilots; to learn more aboutscientifically objective cockpit workload measurement; and to develop asubjective pilot workload rating scale which includes air traffic controlworkload.

(VU-GRAPH -- CDTI)

Another part of this effort is the work to determine the benefits andliabilities of Cockpit Displays of Traffic Information, and to providevalid information to regulatory elements of FAA on the capabilities andlimitations of Head-Up displays.

(VU-GRAPH -- NADIN)

o Efforts are underway to improve the vast FAA communications system. Workis underway to characterize a National Data Interchange Network (NADIN),which will provide improved and more cost-effective interchange of FAA'sdata. This is a critical element in the modernization of the computersystem and should permit efficient and cost-effective integration of avariety of services, ranging from the automated flight service stations tocommunications with remote locations for maintenance data processing.Voice communications programs include the development of modernizedcommunications voice switching systems for air traffic control.

(VU-GRAPH -- MICROWAVE LANDING SYSTEM)VU-GRAPH -- FLIGHT SERVICE STATIONS)

o The completion of the development of the Microwave Landing System i-proceeding, as is the development of automated fight service stations tobetter serve general aviation.

(VU-GRAPH -- FULL SCALE CRASH TEST)

o There is a major FAA effort on aircraft safety, including fire safety ,transport safety, aviation security, and the specific safety problems ofgeneral aviation.

(VU-GRAPH -- HELICOPTERS)

o An important part of FAA's research and development activity concernshelicopters. We are striving to provide for efficient helicopter IFRoperations and are working on a variety of services for helicopters andthe examination of problem areas, such as helicopter icing.

L6-

o Studies of separation standards with a view to maintaining and assuringsafety in an evolving system concerns more than on- and near-airportseparation standards. Work is proceeding to examine present standards andto lay the groundwork for improved standards for vertical and lateralseparation. Of particular interest to the industry is work we hope to getunderway in FY-1981 on reduction of vertical separation from 2,000 ft.above Flight Level 290.

o Separation standards are also a significant part of FAA's Over-OceanSystem Improvement Study, in which an international body is examining therequirements for future operations over oceans and in areas where ATCservices are limited.

o A major effort is underway to examine the navi ation system of the futurein terms of the requirements of aviation. A DOT/DOD activity is underwaywhich is intended to pave the way for decisionmaking in the late 1982 timeframe on the best future navigation capability for the United States. FAAis an active participant and has a series of programs related particularlyto establish the present and future capabilities of systems like Loran-C,Omega, the Global Positioning System, and self-contained navigationsystems.

0 Closely related to this issue is the question of satellite applications.Satellites remain candidates for operations over oceans and are beingexamined for that purpose. In addition, efforts are underway to studyonce again the application of satellite services for aeronautical fixedcommunications to determine if and when services can be competitive withmore conventional land services. Part of this effort is examination ofthe DABS system to determine whether a version of DABS can be used in asatellite surveillance system--when and if such systems become costeffective.

In this brief overview, I have not covered each of the programs, but havetried to give you a picture of what concerns us, and efforts which FAA mustpress actively to achieve results.

E0

CL!0. CU

00ccs

UC)

o 0

LL w.

0 cnrCC) 7 4)

C)' Ci) a)CT) Q) (I

>>

ca C

0,,m) CO L

(~Jn QC:

Lu U 0' C 73

C) o

IU)

UCD C13 0

rrl 0 C'i o- 0- C-

o 00 0

In ~ 0 C)-

a))0r - NCD~)= -4.

-C

<L

QCJ) E 0

H m c

E) C) C

CD ) p 0 11

I-.b. C) 0>-LU Go t- 0 ~ ~ 0r

0 C) ,.C

CLi C) Q Jr

Erc CO)~ 0 al~ C)- .

+ cc CU~ Cc~

< 0 C.)<

0O 0 (S

Mt E0 0)

Cc-Z Cio -3-0. rn

.1- jcu

> LU >-~

(Ii'

4-J -Jf . C . -

-Z5 0 c-i0 w >LL >>

N cc

4-a 4-J

co C3 0) C CCD. 0 4-f a

6U E V.C/)

ocU~ 0

O~CU

0 .0 o ~- E 20

>. CU cn U- 0

o!- 0 CU - CUV VC c

C CL

a CL

0 >.C CC0 E w, E C

0 o 4) 0

.2w0

Z.1 w 0 -c CL~ Ca.2Co .5 CD-~

C m E =, 0a, Co (M. c

e* % Co 0

(jol >4-C (0. o o

-E CU 0.c L

0 0)) 0C4- w - +# U o

0) CC E

cc EEJ

CL E a 0 C~ 0 +

L. . L

0)~ 0EO 00)cC0C 0O C, >

*0 0 000 0 0@

lp

A

b r

, 1-ft

A i! -Apell

77

-7, ccj

C c

"I VM.,I F QJi , rmll%

vo ;,

4k Ju

LI)LT

V

~ * 1 Sri

~ v~t'.~ 4:5zA

3, ~*sv.s~:~'

'S

'I -'V

2 -

W

V "

C.4~ -

A.

/II A

* tr 4 p- t..

U 'S

U -.1 4~:r.v p. - ~'~

1 i A

-~A-- ~, )~

I

.1

U.~.. t .'I

* .ip

~CS

w4 -

AC

Cf

i2~tU tB - -eq.,* 4W' )~4 .4U - -

1 4:.J 4

.I~ fl - C

C.)

LL LCD C

t-- /'1

cli cl

-Z

, c o co ) n,'"

•C D 0 c.),

: O N I,7,

.. .' ,. i .--

I , f

I. //. ".,Jc (. .

iV>

i/

/

7

// 7

( 2

I)0 I

I

~ 4..' (1~ _________ I'0 I / ~ ti~ i

~ 1/ ~1 '~1*.-.~ Y~

o IF L I

U..

I/V

H ~9 C/ .

t gI________/

U'

. CL ;

<~ >

0 .0

oQo&; In

Z C03

CIF 7< _i

-0 0.--

CD L

an

II() CID

U73

41 mi

AN.A

Co -

~cu

3\r

00---0

mk'sI

(1

-r-4-

, it

0 O

oc c1 r)

C.,~ Q) ,co. 2.

i z-zu

0 0 CIr

-. ~a)CJ (j) (C)=C) )C

..) C)20 L_ _

0

V) u ) '.)D)

-;; C

0] 0

0 -- +T3Vz

---- ----

C)C

C))

c. 0 -0C

C.,

C)-)

co L CU) 0) 1

L)C

ti-- 0

Ch7

(c r

C) CD +- -0

C)-

CL u LU) '

C.I C) '

C)) V,) L- ( ! 1

cN L 4 -j ()LL-~L-aCC < >~ -. a ., .7 L

C)) c)

CL Cl) a~ 0 " ~

0C) . -- iC4-0 .C) 0

_L c~ c _n

=

E

0C

C. E ' C)

a) 7a-

E~ E C)

C) LlL

0 C

ccc

to co -

C) 0~ ) C~ V r

C)0 C-L .) c ) c

V ~ C)) I-- 0c

Li. C' > 5

4- CD 0 <(A) a. L) -o U) L) C

4-= 0< -: .L7-

CZ) (o 1) C) > C:!.-C <.C Q - > v-' oClo ca 0 C: F- -0

r C n I =1 Cl - .- , Ca) Ca :: C o) C/) (a +j4.Cy.- ~ c , C c C,-0Z>C:

- 0 Q ~

CL) a 0 >o~ C) U

0. 0)< Cl) 0- wucu 0:

w > 6-

IX-

,4 f

0!qlit

* ~IL~tIr~v -

r

- .. 0

-

110e,-. .~..- - . -* - ~ ro

I * '...

- 1~ A.-

lit~ ....

y~.4

CJ

CO)

-IgooUq. -0. 1

7

1~

-~ I~ A- ~ . S

A'~

.$ *J'~i ~i~2 -, '**% I,*. A'' **

.v'~ ~* A-'- *'4 4-

,, V. I *

4: '- *14

" ~ *~ .

N~ *1*

~

j *~4.*.4 -4

~ $

L A *

4 . . &. ~* - *1

~ .' ~( 7 ~. . t .

f e 1 ' ~

4!~

qpP ~r.4 -. ~ A.4.'.,- 'a0L F

' k 12' * 4

I'. / % -

~ ~' ~I I -~ 44 ~. *~$

9 w~ tI~ ~ ~ r~2 ~,1 .~,A&1* ~

a, '.

.4, . __ A

-A.,' t~I~4

~ ____ ..~

'_____ b4 4

r

pI xkI j

~ 47

1 'jr

4W, AI

~Z V. "

rf -- r- t *

0 .M2~

Us ', At

), I

or

4,V,

lp

L06

4 '%

rt4t 74

.1 AiIt

J

If

prOIL

71.

Vt

1 1 4

L

IL .1

II*4 Lit

SCENARIO FOR THE FUTURE SYSTEM

bySiegbert B. Poritzky

Director, Office of Systems Engineering Management

It is important to understand where FAA's E&D prograT, is leading. What can weexpect? What development and implementation must be started now to meet thedemands of what we continue to believe will be a growing system?

FAA has Just completed an ef ort called, "New Engreering nd DevelopmentInitiatives--Policy and Techr ology Choices," in which we invite all elementsof the aviation community Lo Join us in examining t'e prob',-irs of the present

and the future, and to offer candid recommenda", thit w;1y to go fromhere. We found this process valuable, not becau:, .-,,vnt,'ar new ideasemerged, but because both FAA and the aviation rr:,, , &ane< a betterunderstanding of each other's problems. The coninun,', 2t~v .&A a series ofinsights and recommendations which ae 'nf'uencino,, , , , -nt e'for, othe future.

Using the recommendations from the ujser comTr., " ,i w-rk, FAA hasundertaken to postulate the system a we be! ev -, ' m , k,'d work aroundthe turn-of-the-century. FAA has lookeil at tt- Or-: ", an Operat ions

vantage point, as well as an assessment of ipec. ' " 3nd the possi-ble evolution of aircraft and )vi 3t ;on cnmmr,". ' .

I would like to briefly discuss a .,ulir-T y - f Ih nnr .nr will

undoubtedly change, but we thin;, that tni, w,,-k A I p , cret. , roadmapof development and implementation to achieve ,,- :ap.) : ,e. e feoy will be

essentia' for the future. Let me begin withr un prt:- pts ' the future ATC

system design.

o No change of the system h 1 be perm rt-l ' J - . et . On the

contrary, new systems or procedure,, wi , , 1 iigtl ink reas, in

safety to be gained.

Changes to the system will occur in a, 0vr)" jY'" whenever posi-

ble. However, as system neerl change in h, ' no derision can

necessarily be permanent or final. ,, i rnt 1 ity mu,'t h'maintained to accommodate revolutionary ch.inq-s, If , ..

o Traffic levels will rise at approximateIy th,. r ,, t y -- ')t- c'd levels.

No artificial constraints will be assumed, iv,' if thtre is less growththan predicted due to rising fuel costs, in1cses in basic system

capacity must be achieved to insure the absnljtc. : n mum of system delays.

o A very high level of protection against midair collisions must be provided

by the FAA system. The techniques for providing this protection, as wellas aircraft equipage requirements, will vary depending upon the nature anddensity of air traffic activity. However, in order to maintain the

highest level of freedom of the airspace, "see-ano-be-seen" operationsunder visual meteorological conditions will t1e available where possible.

o In some airspace, services will be provided by an essentially fu7lyautomatic ground-based system coupled with airborne automation systems,

and will be the only services available. However, all properly equippedusers will have access to the system.

o There will be airspace where no ATC services are provided and none arerequired.

0 Automation of the Air Traffic Control system, will te pursued wherever it

increases safety, productivity or efficiency.

o There may be shifts in the balance of c~ntro- responsibility betweenpilots and controllers.

o Only relatively little growth in the number of maior airports can be

assumed at this time, though it is recogrize, trat major increases in

capacity can only be achieved by addition of new runways and new airports,

and by raising IFR capacity to VFR capacity !eve>. Continued emphasis onFAA's satellite airport program will offer car> 'iy increases, but will not be a substitute for new runways and new 3irport.

o We perceive an ever-changing mix of air traver:, ircliuoirg a significant

rise in the number of commuter operations, genero aviation, helicopters,and business aviation aircraft.

o The Air Traffic Control system in the yeai ?uuO and beyond will continue

to be an integrated system that accommodates both civilian aircraft a, Imilitary air traffic.

o Energy conservation will become an increasingly cr'tical factor impacting

the design of various Air Traffic Control programs.

o Finally, provisions for special use airspace . continue.

In postulating the system of the future, we have ni: o make a series of

assumptions which are based on our best current tninking:

o One is that a great deal more automation of th(e air traffic contr-lprocess can be achieved, and that our computer systems and their safety

back-ups can be designed in such a way that automated decisionmaking andautomated clearance generation will be possible and acceptable. We havelearned a lot since the romantic notions of the 1950's that automation wasaround the corner and that the whole process would become automatic. Yet,

if we are to provide a high level of flexibility in aircraft operationsand permit conservative fuel use, we believe the task of air traffic

control--that of traffic separation and efficient flow management--willrequire juggling more variables than can be done successfully by humancontroller teams alone.

o Our second assumption is that it may be pructical to pace more respon-sibility into the cockpit, since technology advances and a betterunderstanding of human factor- may offer the poterit a' for redistributi,onof certain responsibilities between the ground and airborne portions ofthe system. These changes may involve additional pilot responsibility forseparation in addition to the navigation and cicarance execution functions,without additional total workload. They may also i:illude airborne device.which permit a higher level of self-separation in certair, airspace, auto-matically coordinated with the ground-based air t-affic control separationservice.

0 Third, we have assumed that the primary air traffic contro' sJveillancesystem of the future will be toe Discrete Address Beacon System (DABS),and that while the navigation system is not lke,;to criange dramaticallyin coming years, especially in a domestic environment, new systems such as,the Global Positioning System (GPS) may come into Q. n <ome airspace.

o Users have an intrinsic right of access to the sy te: _imitations orrestrictions will be imposed only when no other -rjc,-ur'so i 3va~lable.They will be -emoved as soon as possible.

o Finally, we are mindful of the potential t ,-act ios t i o: asatellite services and "space radars." Th,,, new t. oo, s c rF. beacconnodated in the system when they prove capah'Y, ard c'et-effecl ive.

So much for assumptions of future capability.

We believe there will continue to be several 1,o.euIs :)f a~r traffic cont oservice in the United States, ranging from a highly automat.ei system servngfully participating aircraft in certain airspace, il lin way to a!rspacp inwhich no separation service is provided.

The basic air traffic control separation servie vV.i I rc Cnt inuo to be .ground-based service based on knowledge of c,:-rent i-c(a+t position i nJintent. Ground control and flow management will remaioi impG tant in e routeairs pace to assure fuel-efficient flight paths and t handle rerouting due toweather. It will remain critical in major ter,inal ireas and in adjacenttransition airspace to optimize airport efficier.zy ai, aircraft fuelefficiency, while protecting against collisions.

In the first level of service, applicable to airspace whe there is f1llparticipation in ATC services and in major terminal areas, tn, highest levelof automation will be employed to support the separation service. The controlsystem will be designed to offer very high reliability and reversion capa-bility to a safe, back-up automatic control systor. Re'iability nf theautomated system will be achieved by top-down redundancy at individual controlfacilities and facility-to-facilvty backup using d f'exible data communica-tions network. The automated system will be birked-up automatically by theDABS/ATARS system which will independently provide information directly to thecockpit and to controllers. Another level of protection, which may also serveto build confidence in automation, may be achieved by ise of cockpit traffi-displays utilizing the DABS data link and/or an 'ndependent BCAS. Flexible

use of area navigation and an automated ground-basel system capabie ofhandling large numbers of variables will allow a high dtgree of routingfreedom, both laterally and vertically. Our work on Automat~c En Route ATCsystems, called AERA, will serve as the basis for this system. We believethat even in this airspace, some degree of self-separation may be possible iftraffic information can be provided to the cockpit in such a way that aircraftare aware moment-to-moment of the position and perhaps the intent of othe,aircraft, and that procedures can be worked out to assure that pilots' andcontrollers' actions are fully understood and coordinated.

In the second level of service, provided in low and medium-density traf~i,areas with mixed IFR/VFR operations, the service will be similar to th,0present system, but with automated tools such as metering, sequencing,conflict alert, and conflict resolution advisories assisting the controller.All aircraft in this service level, as in the first service level, will carryaltitude-reporting DABS transponders to permit DABS/ATARS and BCAS to operateeffectively with the ground-based separation service. Although IFR aircraftequipped with suitable area navigation systems will be less constrained thanin today's system, route freedom is likely to be more limited than in a moreautomated environment. Self-separation by pilots, with the knowledge of thecontroller, may be available to aircraft equipped with suitable traffic infor-mation displays, but direct ground control of the system is likely to berequired in terminal areas.

In the third level of service, 1FR operations under instrument MeteorologicalConditions (IMC) will be permitted on the basis of rocedural rules n 3manner similar to that currently employed under non-radar procedures. Thisservice will be available to provide some degree of control to protect air-craft in areas where there is a limited number of aircraft on VFR f]ightplans. Depending on the level of collision risk to public transportationaircraft on flight plans, aircraft will carry altitude-reporting DABS trans-ponders to allow BCAS or automatic DABS/ATARS services in areas where surveil-lance service is available to provide separation. Se~f-separation may becommon in this aiispace in IMC to the extent tnat it can be achieved throughthe use of on-board or ground-based surveillance, a data link, and display oftraffic of concern in equipped aircraft.

RNAV-equipped aircraft will have great freedom in tLe choice of routes foren route navigation, limited only by airspace protection requirements foraircraft on IFR flight plans and restricted airspace.

To some extent, and in certain geographic areas, there is an increase infreedom of choice in selecting air routes today. An increa~e in this freedomcan be expected to coincide with implementation of the programs discussedabove.

In the fourth level of service, no separation services would be provided byATC, but airborne collision avoidance systems may offer a level of protectionin addition to "see-and-be-seen."

Perhaps the most precious aviation resource is the airport itself. Sovera'major airports are nearing saturation, and others are expected to approach

saturation in coming years. It is difficult to predict when a significantnumber of new major airports will be in operation in the United States, andhere we come up against a very difficult problem. While we have made impor-tant strides in increasing airport and terminal area capacity, there are nomiracles waiting in the wings which will dramatically increase capacity of ourexisting airports. We must anticipate that the demand wi2l outstrip thecapabilities of the technology.

Real-time knowledge of actual airport capacity and use of automated runwayconfiguration management systems will permit the use of optimum airportconfigurations and scheduling of arrivals and departures to achieve limitedcapacity gains. Better control of the runway itself through reduction ofrunway occupancy time will allow certain capacity increases. The wake vortexproblem will continue to affect airport capacity.

Reduced interarrival spacings may be achieved by reducing deviation in tnetime of arrival over threshold--probably by extensive use of airborne fightmanagement systems, by wake vortex alleviation schemes or full-time workableground-based wake vortex sensing and avoidance systems if those can beachieved, by real-time runway occupancy monitoring and control, airportsurf i surveillance, and possibly runway/taxiway guidance and cont-ol systemsat t . largest airports. An important caution here: Ten years of wo-k havpnot yet produced a workable solution to the wake vurtex problem, in, itremains a serious brake on capacity. Work on this problem must continue.

A variety of remedies such as the use of short, staggered and dua*-'anemultiple runway combinations, along with carefully defined precision approachpaths and missed approach paths, will be necessary to minimize wake vorteximpact, to accommodate growing traffic, and to permit separate and independentoperations from long and short runways at major airports. The use ofappropriate airport surface surveillance, area navigation computers, theMicrowave Landing System, and perhaps cockpit traffic displays will permit theuse of independent closely spaced parallel runways, possibly to a spacing of2500 feet.

A significant element in attaining optimum airport capacity to major termina'complexes will be a flow management system which, starting in en routeairspace, will provide for optimum feeding of traffic in the most expeditiousfuel-effective manner. The flow management system, which integrates thefunctions of national flow management, en route metering and terminal "low.will use automation tools to accommodate the best possible integration of avariety of capabilities. These will include optimum fuel-efficient flightpaths, the capability of 3D and 4D area navigation, adequate wake vortexprotection and an optimum metering, sequencing and spacing system.

Long-term information regarding weather systems and terminal and capacityinformation will emanate from many places for national coordination. The flo%management service will not evolve into a closed-loop automatic control systemincluding aircraft management or throttle management, but will evolve into onewhich provides automatic instructions to aircraft which can be executed eitherautomatically or manually in the aircraft to maintain optimum runway feedingwhile allowing maximum freedom of flight profile selection to pilots ofproperly equipped aircraft.

' 2

We think that a 3D or 4D navigation management computer wi'l become the heartof many aircraft navigation systems. This wil' make navigation less route-oriented and will let the navigation computer select different navigationsensors for various phases cf flight. VOR/DME and perhaps increased use ofDME/DME will remain basic to the system for many years, although we look toincreased use of other navigation services. We expect tc see major improve-ments in inertial navigation capabilities with important reductions in thecost of ownership, and we expect to see continued use of OMEGA, some limiteduse of Loran-C, and possibly NAVSTAR/GPS if the goals for that system can beattained. We are currently engaged in an extensive study to examine thenavigation system of the future.

The Microwave Landing System will substantially replace the current InstrumentLanding System. Precision departure and missed approach paths will be in wideuse to achieve optimum departure capacity, anu curved and segmented approacheswill be used for efficient operations, noise control and selection of pathsfree from the effects of wake vortices.

There will be important changes and improvements in the weather service. Thebulk of weather information used in air traffic control will be obtained froma National weather network and fed to FAA facilities. Primary radar will beused predominantly as a source of weather information. Some air-derivedweather data will also be fed automatically to FAA facilities for processingand distribution. Substantially improved weather detection capabilities inFAA terminal radars will be explored. We hope and expect that the applicationof new technologies such as Doppler weather radar and laser technologies willenable the provision of terminal data concerning severe weather, accuratewinds aloft, wind shear frontal movement, wake vortices, visibility, cloudcover and other atmospheric phenomena. Although services will continue to beavailable in flight through voice channels, weather information wi7l berelayed automatically to aircraft equipped with data link terminals anjdisplays.

The improvement of operations over oceans is Lhe aim of an international goupwhich is reviewing the application of satellites and other techniques to civilaviation. This group is currently in the process of examining improvementoptions based on an extensive study of current and anticipated traffic.

Safety and efficiency are the primary considerations in this work. The

following are improvement options which appear to show the highest promise:

o Reduction in vertical separation minima above Flight Level 290.

o Reduction of communications delays and application of automatic dependentsurveillance and direct pilot/ controller communica-ions.

o A separation assurance device, such as an Active BCAS, may be an efficient"tail cutter" of the tails of position deviation distribution in rela-tively high-density over-ocean traffic, and serve as a safety aid in area'of the world where air traffic control services are not we1l developed.

o Improvements in inter-facility communications and optimiza'on ofprocedures.

AD-AIOO 370 FEDERAL AVIATION ADMINISTRATION wASHINGTON DC F/6 17/7

UNLSIID THE FAA PLANS AND PROGRAMS FOR THE FUTURE AIPORT AND AIR TRAFF-ETC(U)

22V 80 mmm iEECLASEIFEEEEEE

THE ROADMAP TO THE SYSTEM OF fHE FUTURE

bySiegbert B. Poritzky

Director, Office of Systems Engineering Management

(VU-GRAPH #1 - CAPABILITIES OF THE ATC SYSTEMAl IHE TURN OF THE CENTURY)

I would like now to discuss the roadmap, or transition, from where we aretoday and what we are doing today to the situation in which the servicesdescribed in this scenario are available. Let's look at a top-levul summaryof the expected capabilities of the ATC system of the turn of the century, andkeep that in mind as we translate it into the services which are implied by it.

(VU-GRAPH #2 - ATC SYSTEM SERVICES)

Let's look at the services which are required to achieve these capabilities.A series of services are now provided by the air traffic control system.These are at the top of the list. Added to it are additional or new serviceswhich are either in development or which will need to be developed in order toachieve the capabilities implied by the Year 2000 scenario. Not all of theseare certain, and several will require future assessments and future decisions.

These services will require the development and then implementation of toolsto accomplish the Job.

(VU-GRAPH #3 - THE TOOLS)

Once again, the vu-graph shows at the top the tools which are currently inuse. Then the tools which must be developed and implemented; and again, thereare questions which must be resolved.

The development of these tools must be completed and their implementation mustproceed if the capabilities described in the scenario are to be achieved.

But that's not enough, because if the system is to be productive andmanpower-efficient, many of these services will need to be integrated andautomated.

(VU-GRAPH #4 - INTEGRATION CONCEPTS AND SYSTEMS)

The vu-graph shows three efforts to create that integration and to createautomation which will permit these tools to be used effectively.

94

o First is a National Integrated Flow Manqement Sy,tem-- n et fp.t, animprovement of existing Central Flow with better weather information,with better information on terminal capabilities and better tools tnmanage and meter the flow where that is essential.

o Second, and in a way a subset of the first, is lerminal Integrated FlowManagement--a concept that pulls together with automation a wholeseries of terminal tools and capabilities--ranging all the way fromoptimum en route and transition flow metering and sequencing throughterminal metering and sequencing, adaptive wake vortex spacing toairport surface management, and including the best balance betweenachieving optimum airport throughput and optimum fuel performance forparticipating aircraft, airport configuration management aids and bestintegration of new aircraft capabilities into the flow system.

0 Finally--AERA--which, in its simplest terms, is automated air trafficcontrol to provide conflict-free direct routings automaticallycommunicated to aircraft. Initially planned for en route operations,it will be an automatic system which accepts and provides flowinformation, incorporates weather data and, for the first time,incorporates substantial automated decisionmaking into the computersystem. It is here that major productivity gains should be achievablp.

These three integrating functions are essential elements in achieving thecapabilities of the future scenario.

(VU-GRAPH #5 - THE FLOW TO THE NEW CAPABILITIES)

We now identified a series of capabilities, services, and tools; andintegration processes to make optimum use of those tools.

It is not hard to see that certain tools represent a core requirement.

(VU-GRAPH #6 - CORE TOOLS REQUIRED)

Perhaps the most important, aside from the massive new computer dataprocessing needs I'II come to in a moment, ar"- the Discrete Address BeaconSystem and its data link, the elements of the improved Weather service, andimproved and integrated voice and data communications systems. Without them,the prospects for achieving the desired capabilities evaporate.

But there is perhaps an even more important bottom line.

(VU-GRAPH #7 - CURRENT ATC SYSTEM SERVICES

INVOLVING COMPUTER UAIA PROCESSING)

Looking back at the services, recall that the ones shown on top representessentially the total capability of the existing en route and termini.

S)

computer systems--the 9020 and ARTS systems. Adding even very limitedadditional services to those computers will be extremely difficult, if notimpossible. Some of the tools can be developed by external add-onprocessors--but our studies show that this is not only very difficult, butlikely to be extremely expensive. It is certain that many of the functionswill require a new, highly capable computer system for both en route andterminal applications.

(VU-GRAPH #8 - THE ROLE OF THE NEW COMPUTER SYSTEM)

The requirement is therefore for a replacement of the 9020 and the ARTSsystems, architecturally capable of evolving to the full services described inthe scenario, with sufficient flexibility to permit later options to beexercised with respect to future new communications and surveillance systems,changes in philosophy on self-separation, and a number of other capabilities.The new system must have a level of reliability and availability and aninternal structure to accept the AERA functions as they will evolve and theautomation elements of the Integrated Flow Management system in a way that wecan be confident that real automation of the system--automated decisionmakingfunctions--can be achieved.

If we do not proceed smartly with the development of that new, more capablecomputer system, once again the possibility of achieving the capabilities inthe scenario evaporates.

(VU-GRAPH #9 - A ROUGH TIMESCALE)

There is finally the question of timing of the completion of development ofthe tools to achieve the capabilities we have been discussing. We are at workto try to build the details of the roadmap to show the way the timing mustinterlock, if the developments are to come together wisely for implementation.

On this vu-graph, we have shown time bands which represent the time we hope tohave initial operational capability of the necessary tools. This is apreliminary assessment, which will be refined over the next several months.

It points up the amount of work to be done, and the need to move rapidly on anumber of the major program efforts.

A good many of these dates depend on the successful outcome of research anddevelopment--some of which probes new territory. It depends also on steadyand consistent funding for R&D activities. It depends finally on decisionsand choices yet to be made in several important areas, and that brings us tothe discussion of alternative approaches and their impact on the developmentof the future system.

96~'

E 0

00

co LiN

3~ c-

U)~

Co 0) 0CD m

mI ED '4-coF < :

>! .0 o , ) C -,2 -

co+) - 0 c CDC0 MD 0 E

5 0 0

.2 c- m - c-O

cn C=.N0 <, CL 0 )L

M0) 0 > 0 CDa) -- L Em z: C CD_

1-. L~E E 0 a ."a ~ m 0 Cl

CL 0 CD 0 C4 CLa)CD0 Z 0 0 c

u_ u- COC co L) 0

_L LL

c

E-

0

a)014> C :-

c 0) C

*~~ 0)l c)-

Cn (D t:o a C U)(E 0 E CD C~Of

4) *0 0. E Cc0 cmi2-(. C. C C C.~~~~~ C) U2) Vm ~ o ~ E

M 0 0) LuW~ 0)

I'm c 0 0c)

0 4 . - . 0 ) C 0 C 0 ) 0 U V ) Z0 L O.- cn CCV ~ o

'0 0 6s 0 ca, (n0< Cct tEon 2( < )E

=-4 > w u- 04 -. I C= -

u 0

0

> co C (nwi wf Z 0 ca

a. C < dU 0 02 < V)

M wU%Ni -a-

E cE < )WE m 0E% E i

>-Cc 3 i 0, .Eno c)0-0 ~ Zo (n -~

E E o B E 0 E~

Q)~ .o C4C _

cm C~U)~ Od0 i 0E~ dic 0

t0 di di 0. 0cd

di~~~ > ~ -~~ 0 i~' d~ j

~ m inU. 0..I J Q 0 0 EWCLZ>.w U.< . ~ o z ~ o U )LELc 0

EE

ccCI Ej zC)u

Z Y 0) < 0Ct

c ZC C CI

2o I__

I _ _ _ _ _ _

L ois' siea

.4,I

.~ ~'' ; .1~I~1 -,

~! ~

SQ

q) ______________________________

~j ~ ~ ~ ~ iI ~

- ~ '.

~ * %~ ~ -~ V~- z~ ~

S -.~ 2 ?~ 4;0 '1

I'm _

*1 :1

~iim.m b~

j1--..V.2 V

I 0 1

Hit -

c- c E >M

a '~ bE za)

0?0

0

E

00

E~a 00 L>~w~

0~ 0

*;

- 0 -

U) 4) o_U, C U)U-

E2)CE C. CO

0 = - M~0

(m OutmM C cOL)0 C m 0 n c

E C

- 3

U ~ .1

~ ~ ~

As~m _________________________________________________ -~

eQ

I~IE i~ijC.)

kIm

71 3 - -cI I

.1 ~ i ~ 4a C

Id~ _

1 j~ I '''nhA

I 04

0U0 00c

Q) 0 )

C)

C

00E

CD 0)Z

0 ..0--

c a) E

E , cz ~.u E 76E -5 .C Q) t:_

>a) c 0 )

c~~~~~ E- .9 c ) a .;-aL) CC >a ~

E u~~ C 'Z 0 c~C ~-m0 >

- ~ .- c -. w a<~ > c-

m0 a, 0n 0_ cuCL 0 V z ::C) W

MQ aa

r ' L CD )0 0 2 5 4 0 -0 -V. . ' w <- 0) - C W

THE IMPACT OF ALTERNATIVE APPROACHES iNAIR TRAFFIC CONTROL SYSTEM EVOLUTION

bySiegbert B. Poritzky

Director, Office of Systems Engineering Management

One of the challenges in developing a scenario and roadmap to the future airtraffic control system is that there are philosophical and technicalalternatives which need to be examined and resolved and which can make asignificant difference on the road to the system of the future. In a sensethey are forks in the road, and we must have the foresight to assure theydon't turn into roadblocks. Let me cite a few of them:

0 An issue long discussed is the potential for and value of changes inthe balance of responsibility between the pilot and the control systemin the operation of ATC. Views range widely, from concepts which woulddepend very heavily on aircraft self-separation based on the carriageof airborne devices to a fully strategic, fully ground-controlled airtraffic control system. Terms like "distributed management" and"strategic control" are widely used, but seldom well defined. The factis that the air traffic control system has been and remains adistributed management system in which both the control system and theaircraft crew have specific designated responsibilities. The realquestion is whether there are beneficial ways to change that balance ofresponsibility. FAA is exploring several possibilities in the programto determine the capabilities and limitations of Cockpit Displays ofTraffic Information (CDTI) and in the Alternate Separation Conceptsactivity which was strongly recommended during the New Engineering andDevelopment Initiatives work.

Closely tied to this area is the very old question concerning mixedairspace--the assurance of safe separation between controlled aircraftand all other aircraft, and the role of collision prevention techniquessuch as conflict alert, conflict resolution, and ATARS as prospectivetools to improve the mixed airspace situation.

Much of the general aviation community would like to operate safely butwithout being fully involved in the air traffic control process. FAAis dedicated to preserve freedom of flight to the greatest extentpossible consistent with safety. The direction we will head, basedupon solid data, will make a difference to the evolution of the airtraffic control system.

0 One of the important crossroads will be in the evolution of air trafficcontrol surveillance. All of us see satellite surveillance in thefuture, but we do not yet see cost-effective ways of utilizing it. Forthe nearer term, the issue revolves around the future of primary radarand secondary radar combined for en route operations versus fulldependence on the secondary surveillance radar system-ATCRBStransitioning to DABS. There is a significant cost impact in the

106

decision, as well as an impact on the computer requirement for the airtraffic control system. In the computer development we are planning toprovide sufficient flexibility to permit later decisions with respectto the primary/secondary radar issue, and we plan to make provisions toaccommodate satel te surveillance systems when and if they becomecost-effective.

o One issue concerns the number of air traffic control centers in thefuture, a very old question, and much studied. In our thinking so far,we have concluded that our approach should be to utilize the existingstructure of centers, but to consider placing specific functions suchas, perhaps, high-altitude control and, of course, flow managementfunctions at a few centralized locations. Again, the prospect ofsatellite communications has an impact, but so do safety and backupconsiderations, and the important social impact of major centerconsolidations. This is a very old issue, but one which must be keptin front of us continually.

o The question of the navigation system of the future presents anotherissue. In an activity sponsored jointly by the Department ofTransportation and the Department of Defense, with strong participationfrom FAA and help from NASA, we are examining alternatives, includingprominently the Global Positioning System. The outcome of thesedeliberations and the decision the Nation makes with respect tonavigation will have an important impact--not so much on the ATC systemitself, but certainly on every user.

0 Decisions on the optimum and acceptable level of automation of the airtraffic control process will have a major impact. We believe that anumber of air traffic control functions can be automated. Our work onthe Automated En Route Air Traffic Control System (AERA) and oursuccessful concept demonstration of Automated Terminal Service clearlyshows that the capability exists.

We have, of course, seen dramatic changes in the automation capabilityin modern aircraft, and the best melding of that automation with airtraffic control automation represents an important challenge. We

foresee increasing use of 3D and 4D area navigation systems and weexpect that such centralized navigation management systems will becomethe standard for the future. One of the important challenges in ourIntegrated Flow Management concept is to achieve the best meld ofaircraft/ground system automation.

Making wise choices on what to automate and in what time frame, andachieving community agreement, is a significant crossroad in systemevolution.

o The voice communication/data communication issue is old, but not yetfully resolved. FAA believes that digital communications for safetyservices, using the DABS data link, will offer a great many benefits.Yet progress in voice recognition systems, the desire of many pilots tocommunicate by voice, and the universal view that voice communicationsmust be available for emergency purpose make this a potentialcrossroads issue.

i 0

o There are major social issues that relate to the evolution of the air

traffic control system. In our thinking about the future, we wish toaccommodate an expanding and healthy aviation system in which thepractical capability of the system is not a constraint on growth. Yetwe all know that the attitude toward construction of airports, the

noise and environment problem, and predominantly the funding ofimprovements for the National Airspace System are crucial elements insystem evolution. What needs to be done to accommodate the growingsystem is expensive, and lack of timely decisions to invest inEngineering and Development and in implementation of new services andmore efficient systems will have a dramatic effect on the realevolution of the air traffic control system--no matter how sound theplanning.

I)8

AIRPORT CAPACITY INCREASES--OPPORTUNITIES,LIMITATIONS, AND CHOICES

bySiegbert B. Poritzky

Director, Office of Systems Engineering Management

There has been a great deal of discussion of airport capacity anddelay problems, many studies and analyses, and a significant amount ofresearch and development. Today I would like to provide OTA with anupdate.

Before I tell you about what we are doinq and what we might expectfrom it, I'd like to spend a few moments on the present situation, theforecast, and the financing of improvements.

One important change, of course, is the Airline Deregulation Act of1978. While the full impact of deregulation is yet to be measured,there has already been a considerable change in the character of theindustry. Patterns of service, route structures, and equipment usageare changing dramatically.

Our most recent FAA forecast indicates that while we expect aircarrier revenue passenger miles to increase at a 4.8% annual rateduring the 1980 to 1992 period, we expect only a 21% increase in aircarrier aircraft operations over the same time period. Commutercarriers will carry 13.8 million passengers in 1980, or 4.5% of allfare paying passengers in scheduled airlines service. By 1992 thesecarriers are expected to carry 35 million passengers and account for6.8% of all passenger enplanements.

By 1992, air taxi and commuter aircraft operations are expected to

more than double the 1979 estimated traffic of 4.4 millionoperations.

Business use of general aviation is reflected in the changingcharacter of the fleet. The more expensive and sophisticated turbinepowered part of the fixed wing fleet will grow by 143% from 1979 to1992 while the total fleet will grow by only 59%.

Change is underway within general aviation as well. The value ofsales in 1980 by general aviation aircraft manufacturers in some casesis exceeding levels achieved in the first half of 1979, but the numberof units sold is below the rate in 1979. The discrepancy is explainedby the still increasing sales of the larger and more sophisticatedaircraft in comparison to smaller aircraft.

109

2

Inflation has remained higher than expected. Fuel costs have soaredand are still rising. At the Federal level, aviation must competemore directly with other modes for a share of the limited resource-available to the entire transportation sector. This is importantbecause substantial capital investments will be needed in the nearfuture for replacement of plant and equipment. In the earlydevelopment of the National Airspace System, most public and privatecapital investment was focused on expansion of existing capabilitiesand addition of new equipment. Now, however, our Federal system isfacing a period that requires a far larger capital investment than thepresent rate of replacement of existing plant and equipment.

Airspace and airport demand are expected to increase significantlyover the next decade. In order to accommodate that demand, it will bpnecessary to increase capital investment in the system--new andimproved airport facilities, navigation and landing aids, and airtraffic control facilities. Unless these capital improvements aremade there is the potential that in some areas of the country thesystem will become saturated, and in order to maintain the currenthigh levels of safety, it may be necessary to impose constraints ondemand. We all believe that FAA's mission is to develop and maintaina safe and expanding system of airports and airways.

While the Administration's budget for airport development facilitiesand equipment is a growth budget, we will not be able to buy as muchas we had hoped if inflation continues at the present rate. It maynot be possible to expand the system as fast as we expect the demandto grow.

What does this portend for airport capacity and delay? It is perhapsuseful Lo look at the findings of the FAA-Industry task forces atChicago, Atlanta, Denver, La Guardia, JFK, San Francisco, Miami andLos Angeles and the lessons they've learned as we understand them.

(VU-GRAPH #1 - IFR PEAK HOUR DEMAND-CAPACITYCOMPARISON AT FIVE AIRPORTS)

IFR demand exceeds IFR capacity today at large hub airports. Theairport analyses indicate delays now reach I hour and more peraircraft operation during IFR peaks and that delays averaged 8 minutesper aircraft operation at those hubs in 1978.

(VU-GRAPH #2 - DELAYS AT FIVE AIRPORTS)

Even if only a modest 2% annual growth in aircraft operations isconsidered, the task forces believe that delays may average 25 minutes

per aircraft by the year 1987. This means that even with moderategrowth the problem is becoming increasingly more severe.

, llllllml/m l I m

3

Considering the airport facility improvements contemplated by the taskforces at the airports mentioned, they found that new runways V'ichallow independent IFR operations would afford the biggest benefits byreducing average delays approximately 50%.

Near term improvements such as ATC procedures and more navigation aidscan provide modest gains except at JFK where average delay reductionsof about 30% may be possible. However, even if all the non-technoloGyimprovements were implemented--and many are controversial at theairports and even within the task forces--,delays can be expected tobe 50% greater by the mid-80's than they are today.

Let me touch now on some of the things FAA is doing and the benefitsthat might be achieved. I have said before, esiecially in the "bottomline" presentation on airport capacity and dela' hich we presented atthe New E&D Initiatives Conference in January of this year, that theamount of capacity increase and delay reduction from technology islimited.

(VU-GRAPH #3 - CURRENT MAJOR E&D PROGRAMS)

When the wake vortex problem was recognized nearly 10 yearr ago, twoefforts were undertaken. One, by NASA, concentrated on the mechanicsand causes of wake vortices, and methods to alleviate them at thesource. These efforts have not reached the stage where either theairframe manufacturers or the users feel implementable wake vortexalleviation systems are achievable. The pessimists, in fact, say weare almost back to square 1 on wake vortex alleviition, even thoughFAA and NASA are both trying hard to make headway.

FAA, for its part, undertook the development of wake vortex detectionand avoidance systems and has been moderately successful incharacterizing wakes and developing meteorolog,ca! means forpredicting the probable location of wake vortices. A system undertest at O'Hare has proven technically workable hut has not been founioperationally acceptable by some of the users.

It's clear that the current promise of full wake vortex detectioni arialleviation systems is less than NASA or FAA had hoped. While theresearch and development work will continue, and there is some hopefor active wake vortex sensors, we must concentrate also on otherapproaches, because the wake vortex problem continues to be a majorconstraint to IFR capacivy.

Automated metering and spacing has been long discussed as a potentialsource of at least evening out capacity at a given airport, and someenthusiasts have felt that capacity might be increased. Yet, the,achievement of automated metering and spacing in implementahbl' form

4 -!

4

has proven elusive. We have come to appreciate the truly remarkablecapability of human controllers to manage terminal air traffic and toachieve efficient airport operations--capacities which may beextremely difficult to duplicate with automation. There are, ofcourse, things that have been done and there are more things that willbe done, especially as the problem of fuel conservation becomes morecritical.

(VU-GRAPH #4 - FAA PROGRAMS IN OPERATION)

Several things have been done:

0 Central flow control has been operating successfully for anumber of years and is improving. It has provided animportant operational element which can be fed into anintegrated flow management system.

0 Fuel advisory procedures have been implemented to absorbsevere delays on the ground with significant fuel savings.

0 Fuel efficent descent, a matter of great interest to airlineoperators who wish to fly optimal approach paths for fuelsavings, has had limited success, because of the largedispersion in interarrival times resulting from the existingATC procedures and airport dynamic variables.

0 En route metering, in a late stage of development, is anotherkey in optimizing flow. Effective planning and coordinationbetween centers and airports where a variety of flow meteringmethods are in effect help achieve smooth coordinated flow oftraffic. This is a key element in the effective utilizationof limited runway resources. This will also expand the useof fuel efficient descents.

Is there more to be done? We think so.

(VU-GRAPH #5 - POTENTIAL FOR REDUCED LONGITUDINALSPACING ON FINAL APPROACH)

One study, Just completed, underlines something we've long known, buthave not shown explicitly, the relationship between longitudinalspacing and runway occupancy time. Reductions in IFR approachlongitudinal spacing may be achievable at high-density hub airports,without higher go-around rates or increased controller or pilotworkloads if a) the wake vortex problem is solved, b) theover-the-threshold delivery accuracy is improved, and c) runwayoccupancy times are reduced. Our study indicated that a 2.5 n. mi.

I

5

longitudinal spacing would require a typical runway occupancy time of50 seconds. Average runway occupancy t;ne today vary from 41 to 63seconds, depending on airport, runway. and carrier.

o We think the potential exists to redujce the average runwaytime to below 50 seconds for most current ruowvys if there ismotivation for such reductions.

0 The study also puts a damper on enthusiasm for long-rangestrategic control concepts with runway threshold timereservations provided far in advance of their use.

We are undertaking a project to learn mere thoul the potential ofrunway occupancy time and control, but motivation, and location ofgates and runway exits are important ingredients to better use ofrunways.

(VU-GRAPH #6 - A ROUGH ESTIMATE OF POTENTIAL SAVINGSFROM IMPROVEMENTS AT SPECITEU AIRPORTS)

Let me touch on several potential improvements which promise highpayoff as seen by the task forces and the FAA. These improvementshave the potential to reduce the cost of delays by millions ofdollars, but fall short of completely eliminating them, particularlyif increased aircraft operations and changes in mix impose greaterdemand on airports in the future. We think that Federal expendituresfor R&D to accomplish them are worthwhile, but they must be viewedonly as ameliorations of the problem. Complete solutions are likelyto be achieved only through mijor capacity increases that involve theconstruction of new runways. In some cases, new airports, and moreintensive management of demand, including the redistribution oftraffic between airports may be required to make the most efficientuse of available capacity.

What are we doing about the technological possibilities? Let mereview some of them.

(VU-GRAPH #7 - CURRENT EFFORTS WITH SITE SPECIFICAPPLICATION-

PROJECT: We have been working with the FAA Great Lakes Region todesign a configuration management system for O'Hare toaid in the selection of best runway configuration tominimize delays. A basic system is currently beingtested off-line at O'Hare. It incorporatesequipment/runway outages, wind and ceiling visibility,demand and Midway airport interactions. We believe therunway configuration management system has a great dealof promise in assisting the tower controllers in keepingairport capacity at peak levels.

6

(VU-GRAPH #8 - INTEGRATED FLOW MANAGEMENT)

PROJECT: A program just getting underway, which we callIntegrated Flow Management, may help achieve optimaltrade-offs between delay, capacity and fuel efficiencyat major terminal complexes in the longer term. Theintegrated flow management concept must integrate thefunctions of National flow management, en routemetering, terminal flow and airport operations. Thisconcept will use automation tools to permit the bestpossible integration of a variety of services andcapabilities, including optimal fuel-efficient flightpaths, the capabilities of 3D and 4D area navigation,adequate wake vortex protection, an optimum metering,sequencing and spacing sytem to ensure minimum timedeviation over the threshold, the capability to provideconflict-free paths which recognize limitations ofweather and shear, and runway occupancy time monitoringand control. We will be looking at how to bestintegrate these capabilities into the system and toestablish the impact on ATC automation planning.

PROJECT: We are actively examining the value of separate shortrunways for general aviation and commuters at major aircarrier airports. A recent study of the top 30 aircarrier airports shows the feasibility of such runwaysat eleven of the thirty. A detailed, site-specificstudy which examined two possible runway configurationsat Denver has just been completed. Results indicate theconcept to be feasible and of substantial benefit. Wewill be working with all of the FAA Regions to explorethe full potential of this concept.

PROJECT: We have developed the means to analyze the IFRapproaches mentioned earlier--such as closely spacedparallels, triple parallels and converging IFRapproaches. We are developing a candidate list ofpromising airport sites to apply this analysis method sothat site-specific feasibility analyses at candidateairports can be conducted.

PROJECT: We are evaluating the applicability of terminal airspacemodels which can complement the airport airfield modelswe have successfully developed. These models may b&useful in analyzing terminal airspace capacity anddelays, and especially in the development of theIntegrated Flow Management concept.

7

PROJECT: To aid in managing airport surface movement of traffic,FAA is testing the new ASDE-3 airport surface radar.This sytem will provide all-weather capability, runwayand taxiway maps, with the reliability needed forhigher-capacity ground and local control inlow-visibility conditions.

PROJECT: We are working on the definition 1or an advanced systemwhich would include not only ASDE, but a trilaterationsurveillance capability. This would provide data blockson the display for aircraft and vehicle movement andperhaps the basis for on-airport guidance and control.

PROJECT: Potentially valuable applications of the MicrowaveLanding System at major airports have surfaced. Weforesee innovative MLS applications, particularlyprecision missed-approach capability for independent anddependent IFR approaches to triple parallels, (VU-GRAPH#9-POTENTIAL MLS APPLICATION) and converging runways;and independent approaches to separate short runways forgeneral aviation or commuters.

(VU-GRAPH #10 - POTENTIAL MLS APPLICATION)

Converging approaches are used extensively under visual approachconditions at Chicago. They are not used when the weather goes below800/2 because the aircraft may not see each other in the event ofsimultaneous missed approaches. MLS guidance could allow highcapacity configurations to be used in IFR operations by providing theability to give precise navigation for missed approaches. In the caseof O'Hare, they represent the difference between perhaps 170operations per hour and 135 operations per hour.

(VU-GRAPH #11 - POTENTIAL MLS APPLICATION)

Use of separate, short runways for general aviation and commutersprovide an operational solution to vortex separation problems and helpcope with the growth of general aviation/commuter operations. Butsuch runways must be located in positions where they can be operatedcompletely independently. Full use of such runways at specificairports may require the definition of triple parallel operations, asat O'Hare, Dallas/Fort Worth or Atlanta. Resolution of airspaceconflicts may be required, as for example at Kennedy where the smallrunway exists but is not routinely being used because of potentialconflicts with other operations. The precision paths available fromMLS should help. MLS may provide solutions to potential obstacleclearance, terrain or siting problems at Denver, where a study wasrecently completed on the feasibility of a short GA runway.

8

MLS may help reduce the 4300 foot parallPl separation requirement forGA and commuter traffic on parallel short runways by providingadditional navigation precision, high glide path angles for extraspacing, and, as above, by providing precision missed approachcapability. MLS would also provide the ability to restructureapproach paths at selected airports to permit segregated approaches toshort runways by general aviation, helicopter and commuter operators.The use of the variable glide paths available with the MLS will permitsmall aircraft to follow "heavies" at a higher glide path angle as ameans to ensure protection from wake turbulence.

What about the longer term programs 4n E&D that , improve capacityand delay? Let me mention just two of them.

(VU-GRAPH #12 - LONGER TERM E&D PROGRAMS)

The work being done in the Automated Enroute Air Traffic Controlsystem, the AERA program, will evolve into terminal automation aswell, and will become an element of the integrated flow managementconcept.

The FAA program to examine the capabilities and limitations of cockpitdisplays of traffic may show that they can be a rart of the futureterminal system.

Before I close, I want to say, before you tell me, that I havp nottalked about demand management. It is not a popular subject, althoughclearly redistribution of air carrier and general aviation traffic canreduce delays. Speed class and weight class segregation or f~owmanagement schemes might well produce overall savings, but means toimplement such disciplines are not yet in sight. This matter is, ofcourse, not a problem which is FAA's alone but involves the wholeairport user community.

To summarize, there is much we have learned about capacity and delay.There have been disappointments, but there is far more that can andmust be done to achieve optimum capability from our existing airports.

(VU-GRAPH #13 - SUMMARY OF AIRPORTIMPROVEMENT PROGRAMS)

Perhaps the most interesting and the most valuable airportcapacity/delay improvements in the long run are direct and indirecttechniques to overcome the wake vortex problems, better innovative useof runways and runway configuration, the use of separate short runwaysat major airports by G/A and commuters, independent and dependent IFRapproaches, applications of the Microwave Landing System, and anintegrated flow management concept. The prospective payoff from

9

technology changes and improvements seems high and seems to us to bevery much worthwhile, but we must not he deluded. The technology canoffer important benefits, but probably not enough to solve theanticipated problems of capacity, demand, and delay, even under modestgrowth assumptions. Technology and operational/economic chnices mustbe implemented hand and hand if we are to meet the demand.

117

~-cc

4 CD C

C O

E 00

CD.)I'L

a) 0

Z. C)-(

c 00 It 0Ca)o 0 0

NCO

-C 00 0 C 0

20 2o0 u 0

a) o _c( co +-Lf4 Jn

(nq CLO~ c C

0*- QQ)C

00 z 0O

0 0

120

0

CL)C" E

0 (CO

cu E zCL) _d M

C.) C0 > 0 ciCfi >--E

> ( 2) rC)-

-> .9Li

>( LP CO) ( C L C3 0U~

>4C cu a< cjV

LU ~ c 0~L ~ C.1 0 ~ C

4-JJ F ...

U)l0 w c +-' 75 (0

00

0()x M

0

t3)

0 LCLL

LLC) 1

(D <DC

_0l 0 Q.

C) )

C n C

ri CD~

C))IJ -cr =>E

-4-4- - 4-D

L W~

CCC27 _ 0 (D LOC -C

0U CL- 07j.C~~Q U)>.C --

D (iC

00 -.-

-U) C) U)

a ) ccLf "U

a: 0 -0 0 C

0 >

4- (D.

E~ ~ D o0Q C cc .0 CD ' .

~) ) C~0~ C) (D- >- om~rn C

- >C)i:;: I I C ,u.- CE C U CCr .- 'I-Wu O:3 L

0 c - r 2C..

0.

ocL. 0 C:)

~ < > co H <> (I> *(

ji

0z 0 00 - cr L

LLLL

0 0 1C!

L.. - L F- 0C)

WI LU LU <a) 0 C)C c ) <

h-c-

7u

LOL

-0 0 7 . c Eo coCauCD) o CD <C-.-

- 4-

0 _ 0 0~

0 0U) CU)_

4 o 0 C 0C (n~3Hm - 5 c >~

Cl~~~~~~ a 0m- ca- -

U)a

C

2sft <l c10 ) 0 4- )

0 CC -C 0 -IL - U oc

Co~ > uC C.)

Co CoQ)

CO

>2 C U ) -.< U) lC rU

J (D- c. 0 > uco o -0 UI co

co 0c - 0-

_0) L

>.C CU CDD

0 ~ 02oM. -- Hn L) (

0 C) 0 0 > 04 0

E00)

CTJLcc

L2 0

U) 0

CD0 U3 2COO

_0coE~~~.C

ci) 0

G.) U) CS)

cu e0 E%0Liu Jc al) .2 (

c~ o

L) ) Go0D 0a

U-~C -> U

a) c: 0 -kC0J 0 oc,-J

- CL C) .co Fn CD-

2U _T _- C. aE

- 0>

IIN

Ccc

0.N

(Z~

0- " -

>00

<2.<

_ 0

ca 0 C 4-J U

±t L) . J , -- -- - --

L4. n (I) -

0)

0 CD L4 -- . 110a

)

0C,,

Z)l

0 4-J

C 0DCIO

Cl)

0 C

LLJ 0

D CU ) CO.0

(T) - I- co(/L U) (-)4-

_ 0Q C

ci

0@

0-

Co:0 07

00

0

<C)

. - : C:c o c

(D 0Co>0 co C~

cri 0

CuU~co C: W

U) 0

() %4-- ( ) o 0C0 _ 0

Df) M

UJ(D0 0- .

00 coU)U W :

0 DCI a L5c

(CL 0 .0 .!: 0)-CLl

(DD

Q) c od E I

4-2 ccoU )00_ -:>- --<

0 0 000 L 0°U

) 0 c a) 0

C) _.

- 0 'F -C- >. (_ "-

4-i Mj

0 0' o F-' u -' ..

co o 6-6

'4- .t,4-_- (.

.. ~~U o, o !i ,; :.

0 -- <W

m (D 7.

-00-- ) C__a

00

c4 , I I I -:-- I; 0 "F) C) c)'-

> Ow - E CL C C_

Lam-)

() ':-({,,

U) OxC:

w

C) cn

£1o E 0co

Q) <

co

C,)

C) _ )

- C ) u%+-- w> 0 ) < cu

0 - 0 c )~C1): C U

c ) C) cu

0) L- C.) 0~C

0 C L C C ) _ -

0 _

C#)o

NEAR TERM SYSTEM IMPHUVLMI N 1'j

h y

Martifl 1. o1,-,kyDeputy Director, Sys tems Hesf?;ircl and Hf ,vi, I opvvi, :

I am grateful for the opportunity to speak to you today auout ;:un,the Federal Aviation Administration's research and developmuntL acti. ii..I will briefly describe some of our major activities, including our i :uo:

Separation Assurance Program, Weather Program, Flight Service StatiorProgram, and Automation Program Evolution.

Our Airborne Separation Assurance (ASA) Program has, as it- pri2cIpDlobjective, the enhancement of the safety of air travel by redu(ing tepotential for mid-air collisions. We are actively pursuing two lntercelcitelsystems which will fulfill this need - the Automatic Traffic A/_visury a-iResolution Service (ATARS) and the Beacon Collision Avoidance Systemi(8CAS). The ASA Program elements make use of the DABS surveiliance fnJ DatoLink System.

ATARS is a pilot oriented ground-based collision avoidance systu,.ATARS utilizes Discrete Address Beacon System (DABS) survelilance ()td.,computed traffic and resolution advisories in a totally automated qruunCcomputer system located at the DABS sensor site and delivers theseadvisories to ATARS-equipped aircraft via the DABS ground-air-grouni ci-ta

data link. Once the necessary equipment is installed on the ground, anaircraft must buy a new airborne DABS transponder and a collision avoidancedisplay of his choice. This new transponder can replace his present AT,beacon transponder at a modest cost increment and will utilize texis[Jn5altitude encoder systems (which are also necessary for collision avoijance,protection). It is important to note that once an aircraft has the -A3S ancATARS equipment installed, he will receive immediate protection against anyother aircraft which is equipped with today's ATC transponder and altitudeencoder system and the new DABS transponder. This will insure that tnefirst aircraft that equips will receive substantial protection imweui,,tLy,since the transponder system is already widely implemented and is requirecby law for operating above 12,500 feet and many airport areas.

ATARS will operate In both terminal and en route environments witri,.

DABS coverage, providing pilots of ATARS-equipped aircraft with acomprehensive traffic advisory service and an effective resolution servic,.For uncontrolled aircraft, the traffic advisory service will enhance tht,

pilots see-and-avoid capability while the resolution service will provide,collision avoidance service not previously afforded by the ATC systrm. Orcontrolled aircraft, ATARS will serve as a separation assurance h ar-kip t,

the existing ATC system. ATARS is currently under developmet ;, rd

expected to be implemented along with DABS.

BCAS is a collision avoidance system installed in airreralt a 1,,Fin an independent but coordinated manner with the grounu air triffir, coiOT-ilisystem. A "major" advantage of BCAS is that it travels with the airLratt

thereby providing protection in airspace where ATARS may not hi: depklnYf.u.ATARS and BCAS are complementary components of the FAA's intaren,0,: ,

to provide col ision protect ion fur all ;j ,ii

The current ECAS designs have evolv-, from tihe ,mirum II. Vl_ 1 ii.avoidance system studies of the late 19601's arid the R'A toaisiljiIv

cemonstrations conducted over the 1972-1978 time frame.

The program is divided into two major elements, Active BCAS ano F ji9CAS. The Active BCAS element is a relatively simpler oesicn appropriatefor low to medium density airspace. The Full 3CA', design iw relativeyoophisticateo and incorporates a number of features :ot available2 in 'cLiv(,BCAS for operation in dense airspace.

Recently, Active BCAS experimental uniiits aesigned an fanrlsr ,t-:Lincoln Laboratory of Massachusetts Institue of Technology have beendeliverec to the FAA for flight test and evaluation. In aduition, acontract was awaroed to Dalmo Victor/Textron for industry-oeveloptiprototype units. These units will be installed in air carrier dircru't ti,operational evaluation.

During FY 1981-82, the Lincoln Laboratory Active U'AS units willcomplete tneir engineering tests and will he extensively evAluateu iloperational areas on board FAA aircraft. The Dalmo Victor/Textrnii ilit.,will be delivered to the FAA and their evaluation in air carrier aiwruttmil' Degin. A draft National Standard for Active ECAS was Just pull1r,,anu the development of Minimum Operation Performance Standards for ,rtit,,avionics units will follow.

In its present form, tne Active BCAS system cannot ontain rridirectional information necessary to provide the proximity oisplay tr) 'h-

oilot I spoke of earlier. This is due to the technology nroblems otperfecting a practical airborne antenna which can provide this type L-directional information. However, recent developments in the areaairoorne directional antennas hold promise of liminating this troun].,:restriction in the not-to-oistant future. We are now actively investci<,atir-,,

this possibility ano, if the hoped for performance can be achieve,, :t maywell be possible to obtain air-derived bearing information of modest qualit\

to provide a proximity display with this Active BCAS syst.em.

The second type of BCAS system which we are now pursuing lb udiii_ Li(

"Full BCAS." It provides a more complete service than the Active ECAS animakes use of recently developed techniques and new high speed computU7technology. When operating in radar coverage, this system will proviue

directional information of higher quality than is possible even with "I,airborne directional antenna I spoke of earlier. Because of tWi, i.. ,ifeature, this system can also provide an electronic displiy (,0 'Un¢mIi;,aircraft in addition to both horizontal and vertical maneuver cnimnal,!,, iall environments.

Full BCAS units will be developeo by industry undur a contra t uii(, *'

be awaroeo. Following design definition, engineering models will hefabricateo "or flight test and evaluation by the FAA. This activity n,;, 0r,followed by the fabrication of prototype units for operati r in air ,arri,:aircraft.

AVIATION WEATHER

The overall objective of the Aviation Weather Program is to priv,'timely weather information to all users of the Nation Airspace System,significantly improve the capability for detection of hazardous weatiierphenomena, and to provide rapid access to the national aviation weather aritbase by all users.

The Weather Program is separated into three development areas. Thefirst area is directed to developing capabilities for processing,distributing, and display of weather data to all users in the NationalAirspace System, the FAA or the public. Plans in this first area involvp

the use of the Flight Service Station computers now under procurement toatprovide automation support to the present Center Weather Service Lnitmeterologists in the en route centers. To display tabular weather uata t,'the controllers, the Electronic Tabular Display Subsystem (ETABS) and theTerminal Information Display System (TIDS) will be used. Our NationalAirspace Dioital Interchange Network (NADIN) will be utilized for inter annintra facility weather data communication.

The second area, Weather Radar, involves the development of a newgeneration of weather radar using doppler technology to provide radial wini"velocity and velocity spectrum width in addition to reflectivity data,detection of hazardous weather phenomena is to be greatly improved Lhrokqghthe application of doppler radar technology. A joint program wasestablished in 1979 between the Departments of Commerce, Defense, annTransporation, with a Joint System Program Office (JSPO) established witnilthe Department of Commerce to develop and implemen' the Next GenerationRadar (NEXRAO). The FAA is providing both manpower and funding resources L,the NEXRAD JSPO and expects this new radar system will satisfy itsrequirements for weather radar for the detection of hazardous weather in tneen route enviroment. It should be noted however, that due to sitingrequirements, scan rates and display update rates, the detection anc oispl-ivof hazardous weather phenomena in major terminals may require a separateterminal weather radar.

The NEXRAD program is following the A-109 major system acquisitionprocess, with multiple system definition phase contracts scheduleo for awarnin the fourth quarter of FY-81 and multiple validation phase contract'scheduled for award in the fourth quarter of FY-82. Full scale developmeiLwill be initiated in FY-85 with a production release in late FY-86.

The third area, Automated Weather Observations Systems involves thedevelopment of new weather sensors and the capability to process sensoroutputs into a complete weather observation for voice and digitaltransmission. The FAA is completing the development of two systems forprocessing and generating observations. The first is a simple systemidentified as "WAVE" (Wind and Altimerter Voice Equipment) for provicinq awind and altimeter setting observation with a voice output for use atairports with an instrument approach procedure requiring local altimetersettings; and an Automated Low Cost Weather Observation System (ALWOS); Loprovide a basic observation capability at general aviation airports.

A fully automatic weather obUerv t ion equiva lent to I t .d i*, .:.requires the development of several new key sensor,- in whit'h h)rk t! :H,initiated jointly with the National Wf-ather S ivi i,' iii] t he !iIv t

A joint development and implementation program jetween the Department'.of Commerce, Defense, and Transportation for prucurement of the AutomatedWeather Observation System, it,. expecteo to begin in 1981. io".nt operationrrequirements anc functionsl specifications are scheduled tor approval inFY-81 and contract awards for development and production of semi-automateiweather observation systems and cevelopment of key sensors aie planneo if

Semi-automateo systems wouc be ivi'lahle for fielu aeplr,'ert - n .Key sensor development woilrd se cumpleted in 9t 34n product tl 'urn;ewoula be available for upgradino the semi-automated systems tn fitYautomated systems in 1986.

-L.2HT SERVICE ST-IGINS

Tne existing Flight S ervice Station System is, comprised o! {3stations. Flight services are provided in a highly labor-intensivu ino1n:Nith most services being provideO on a person-to-person bas is. Theprojected growth of general aviation ano the resulting aemaif lir servict:will redjire more than a doubling of sytem capacity by 1995. li meet tnI:-demand by expanding the existing "manual" system would require the hirine ,,t6500 additional flight service specialists with an attendant annual systeoperating cost of $320 million per year. The modernization anG autoImdtiar,)C the Flignt Service System as planned will meet the 1995 demand wito tht:rurrent staff of 5000 specialists or possible even less with an annualnperatinq cost of $140 million/year. The cumulative cost savirgs to 1995due to automation of the system wouiJ De $1.5 billion. Key to this ,JnceFtis pilot self briefing to be discussed a !lit later.

The Flight Service Automation System contract award was male juiie 7,1980 to three contractors, each for a one-year design vcerificationcontract. During this period the contractors will verify their software 6,icomputer system designs. One of the three contractors will be selectec forthe production contract, to produce snd implement FSS automation to twe,steps, model 1 and model 2.

Model 1, limited automation, is to be deployed to 41 level III Fstsbeginning within one year after contract award. It provides at thespecialist's position retrieval and display of weather and aeronautic,.1 ,Jtand flignt plan entry and processing.

Model 2, full automation, is to be deployed to 61 Fts to ne lu.airports which are major centers of general aviation artivity. At tt,,specialist position, in addition to the model 1 capabilities, will nuweather radar and weather graphics and aunlitional aeronautical infurmt iiii,and processing.

The model 2 system has the capacity to handle the long-term fliur,5 tservice demand whether from the specialist or directly from the pilot. Pt

also provides software and interface to support direct user access t(- t-iesystem.

The research and development program ill [-light Service S'tations isprimarily the development of model 3 or pilot self briefing. Widel hasthree major subdivisions:

(I) user self-briefing via computer terminals cilled Direct user AIcct',Terminals (DUATS). An initial complement of DUAgS is incitJ0eu in

the current systems acquisition to implement toe operdtiofiacapability to encourage users to purchase DUIAl tar o;elf-hri,:,and flight plan filing.

(2) Direct user access via push button telephone. his mode ,i :,access requires application of Voice Response system (VRt,)technology and of the capability to automatically proct-< .,-,aeronautical, and flight plan data.

Currently, the automatic processing of data and conversion tw v_,for response to direct user access via push button telphon, it ,,-provides surface observations, terminal forecasts, winds il ,t, ,ncwarnings of severe weather for any of over 700 weath:r r n rt in,:points in the country. This initial capability is ,'rere,,undergoing public demonstration in the Washington, i.., ,. ,the Columbus, Ohio Flight Service Station irea.

As new and/or expanded capabilities are developed, they Wi.i 5,subjected to test and evaluation. Following operation,l urmoii,tt I rwill be scheduled into the sequential updating of the operat 1)il.lI

automation program.

(3) Direct user access via any typt, telephioro,. Iii mO, ,I .,access requires application of Voice ket:,nj)itio n (Vi) te ,The FSS system responds to this input usinq 'hc previiclkdeveloped voice response system. The voice rtotpiitil ., iaccess is currently under test ai evalujti ,jf.

AUTOMATION - EN ROUTE AND TERMINAL

The En Route program includes numerous activities jimei: .it tilt:.'.system, however, I'll only address a few of them here,.

Since the inception of the air traffic control system, tllt Jt !t11,(. 1

posting flight data information to the air traffic contr iltr r c,, l.., *!.paper flight strip. Before the introduction of the present NA' tsystem, flight data information was entertd and updated manal i 'I, .,

on the flight strip. The present system uses electrn-me',.vi, ii I 1,'strip printers which, under computer control, print iriit im ,l tflight data on paper strips and distribute the strJi., th hti,sectors. This system requires mountinq or "stuffi )" t in., Itil, 1, .in the flight strip holders, placing the holders in thie de,ireu prit ii()the flight strip bay, updating the flJqht data by pencil1, ind enteii ,:update information into the computer by meals ,if i manlt iy 1pr~tl........

(ev ice. This i a cumbersume uperatior wih ii cotf v,( (-t i;.: f '"0"

'ontroller's atu -A" rontroller's time.

The Electronic Tabular Display .Ajbsystem (ETAf3S) effort is involv'o withthe application of electronic display technology in the replacement of theflight strip printers and paper flight progress sti-ips now in use at il] AirRoute Traffic Control Centers (ARTG~s). Tnroucp the use of electronicdisplays, processors, and touch entry devices, ETAH-S will automaticallyprovide non-radar flight and control data to the data ("D") controller ateach en route control sector.

At the current air traffic level, approximately 8,000 controllers arerequired to staff the existing sectors in the Air Route Traffic Control,enters. A Recent study indicates that with the forecasteo growth in airtraffic the controller staff will ieed to 6e increaseo significantly withthe creation of new sectors which will Oe necessary to meet future trdffli:

demands. The study shows that the implementation of ETABS will result in a,increase in controller productivity with a consequent reduction in secturstaffing growth rates. Based upon an ETABS implementation in the 1986-1987period, present value cost savings through 1999 are estimateo to be inexcess of $400 million.

ETABS consists of the computers, displays, and message entry resourcesnecessary to support the en route sector positions. A distributeoprocessing network concept is used to accept, process, and distribute flightplanning data to recipient sector positions and to accept, process anatransmit controller originated messaqes to the NAS en route Central Computeromplex (CCC).

The en route metering function organizes airport arrival traffic in rhe

en route airspace by metering flights to their destination airports.Meterec flights are scheduled for delivery to a merge point with a unitorintime separation to match the acceptance rate specified for the merge pointur vertex. The automated metering process extends beyond the indivicualrenter's boundary via existing computer-to-computer interfaces by thee co'ange of metering-related data.

The metering function determines a tentative landing time for anaircraft. A list of these landing times will be displayed for all aircr:iftarriving at a given metered airport.

In order for the metered flights to meet their landing time, delayabsorpticn strategies are generated to absorb any required delay. rlintsare monitored at various times along their routes to determine curreniahsorbabie delay and the appropriate delay absorption strategy, ;uch a;

outer fix, speed reduction, descent or primary hold. These advisOrie, 111dioplayed on the plan view display of the R-controller who either has-mj,,trol of the flight or controls the stwtor containinq the relatted i .

Tre Conflict Resolutior Advisories fiinctiot is desi t2, to pitov i,' l IP,, route radar controller with a dispiay of possiole alternative,' tlilr tl'resolutirn of conf]irt', ldertt i filed h tile conflict alert. tlrret inn. 11prime objective of ,h. Cu r t re-,rjit V i)n function i', to redo, i i,!i

of system error, the violation of en route separation standaros, by reducingdecision-making time in complex encounter situations. By display ofpossible alternatives for the resolution of conflicts, the prospect is thatthe conflict resolution function will offer confirmation to the resolutionpreferred by the controller, who generally can introduce to the situationother considerations beyond the capabilities of the current levels ofautomation, such as weather, communications failures or the presence ofuncontrolled VFR aircraft.

The objectives of our Terminal Automation Program are to provideimprovements to the existing terminal automation systems ano the developmentof new system capabilities which will result in increased safety, improvedcontroller productivity, and reduced fuel consumption of traffic in theterminal control areas.

System improvements for which the development programs have recentlybeen completed include: The field evaluation of an expanded ARTS III systemat Tampa-Sarasota which provides the capability to process digitized datafrom a remote radar site and display it on local radar control consoles anoremote tower cab digital displays; development and operational evaluation atthe FAA technical center of a new Airport Surface Detection Radar (ASDE-3);field evaluation at a medium activity airport of an electro-mechanicalsystem for visual confirmation of voice take-off clearances; andhardware/software changes to ARTS III to enable it to operate with DABS.

Development projects currently being initiated include the expansion ofARTS II processing capacity and the addition of beacon tracking, conflict

alert, and minimum safe altitude warning to this system; development ofTAGS, a capability for interrogation of transponder-equipped aircraft on theairport surface, and development of terminal sequencing and spacing.

A word about terminal sequencing and spacing may be in oroer. TerminalSequencing and Spacing is a development program designed to take the firstmajor step towards automation of a complex air traffic control function:That of sequencing, scheduling, and spacing terminal arrival aircraft.

In general, the sequencing and spacing will control the flow of arrivalaircraft into the terminal area, from the en route feeder fixes to therunway. It determines the final landing order based on each aircraft'snominal landing time, using predefined flight paths, and establish scheduletimes at various points for each aircraft which will assure proper spacingof aircraft at and inside the final approach course intercept. Thesequencing and spacing aids in the control of arrival aircraft by thegeneration of recommended commands to satisfy the established schedules.

An increase in system capacity may be achieved by providing moreconsistent inter-arrival spacing of aircraft, thus assuring an increase inrunway utilization. The automated sequencing and spacing function may alsoprove beneficial in other areas of productivity, that is, an overallreduction in arrival delays and the absorption of any necessary delays inthe most efficient manner available.

Safety will be maintained by assuring that rt-coruneuded control a-tion:,are within the bounds of aircraft performance characteri Sticsf ano byassuring that contrul is in accordance with existingj AIC procedures ifldpractices.

LONG TERM ATC SYSTEM IMPROVEMENTS

byDr. Edmund J. Koenke,

Acting Deputy Director, Officeof Systems Engineering Management

INTRODUCTION

You have heard from previous speakers about many of the activitiesunderway and received a broad perspective of the improvements plann ,for the ATC system. This presentation will deal with some of our longterm system improvements.

The programs I will address are the Integrated Flow Management Progrinaimed at developing a concept to integrate National, en route,terminal and airport fluw management operations such that delays andfuel consumption will be minimized by making best use of airport andairspace resources.

I will talk about our work to automate the en route portion of the ATCsystem, and our efforts to identify new and alternative methods toprovide safe separation. I will discuss briefly our major effort toreplace our current computer system with one capable of satisfying thefuture demands of increased aircraft operations and higher levels ofautomation will also be described. Finally I will discuss our work 'nhuman factors.

INTEGRATED FLOW MANAGEMENT

Integrated Flow Management is a requirements definition, conceptdevelopment, and integration activity. The integrated flow management,concept must integrate the functions of National 'low management, enroute metering, terminal flow and airport operations. This conceptwill use automation tools to permit the best possible integration of avariety of services and capabilities, including optimal fuel-efficientflight paths, the capabilities of 3D and " area navigation, adequatpwake vortex protection, an optimum meterij, sequencing and spacingsystem to ensure minimum time deviation over the threshold, thecapability to provide conflict-free paths which recognize limitationsof weather and wind shear, and runway occupancy time monitorinq indcontrol. We will be looking at how to best integrate thesecapabilities into the system and to establish the impact on ATCautomation planning.

The Integrated Flow Management Program addressps two time frame;,. (Thephase covers the longer term, addressing the total flow managementconcept around the turn of the century. To provide a ,horter termfocus for the flow management program it also addresses the nel, termperiod of approximately the next five to tpn year , and is predir-at-dl

HO.10

upon the assumption that even with time and eq'uipment constraino',improvements can still be made. For example, we have been workingwith the FAA Great Lakes Region to desgn a configuration managementsystem for O'Hare to aid in the selection of best runwayconfigurations to minimize delays. A basic system is currently beingtested off-iine at O'Hare. It incorporates equipment/runway outages,wind and ceiling visibility, demand and Midway airport interactions.We believe the runway configuration management system has a great dealof promise in assisting the tower controllers in keeping airportcapacity at peak levels. We will be delivering a runway configurationmanagement system for realtime operational tests at O'Hare at the endof this fiscal year. We are examining the applicability of this typeof aid to other airports and envision this as a potential element ofthe flow management concept.

A preliminary concept of Terminal Flow Management has been post,1 ate-iand is being examined with respect to its ability to accommodate:

4D RNAVFlight Management ComputersSevere WeatherWind Shear and Wake VortexMetering and SpacingDeparturesMultiple RunwaysMLS ProceduresProfile Descents

and to interface with en route metering.

Concepts for National Flow Management are also being investigated.These concepts are viewed as enhancements of our current Central FlowControl Facility. It is envisioned that this enhanced capability willbe better able to balance the demand on and the capacity of the systemsince it will have better and more timely information and will becapable of directly interacting with all center and terminalfacilities.

The en route element of the flow management concept is containedwithin AERA and we will be addressing that program in detailmomentarily.

These elements of flow management--airport, terminal, en routp andnational--will be combined and integrated and will form the integratodflow management concept.

AUTOMATED EN ROUTE ATC (AERA)

The objective of this program is to develop the concept, functional

Ill '

description, and engineering requirement for an advanced ATK 'v teiTiwhich will automatically perform most en route planning and cont,'oprocesses under the active management of controllers. This sy,tem wl'provide better accommodation of flexible, fuel efficient profils,increase ATC productivity, reduce historic causes of system errors,increase ATC service availability, and reduce the potential for piliterror.

AERA will be a fully automated system, embedded within each en roul-facility, that will (a) automatically plan conflict-free, fuelefficient profiles for aircraft operating in positively controlledairspace, (b) generate ATC messages needed to execute the plannedprofile and assure aircraft separation, and (c) deliver ATC messagesvia a data link or VHF voice channel. AERA will protect agains' systemfailure by providing a coast capability and backup clearances 3nd willbe compatible with the backup systems such as DABS/ATARS and BCAS.

The AERA operational concept assumes that ground-based comput,-r,' willautomatically perform most of the routine ATC planning and controlfunctions under the active management of air traffic controlspecialists. For IFR en route aircraft these functions willautomatically produce a plan of clearances to be issued, eitherautomatically or manually, at appropriate times. The clearance plan,if followed, will ensure that all controlled flights will remainconflict-free, fuel-efficient, and when appropriate, metered.

Aircraft carrying an FAA-accessible data link, area navigationequipment, and a flight management computer will be able to taki, fulladvantage of the capability of the AERA system. Pilots flying ;ircraftwith the full complement of equipment will be able to file for and, inmost cases, receive precise, direct route, fuel optimal flight profiles.

In AERA, the volume of airspace within which the computer can controlflight movements by issuing clearances to pilots is known as its"control region." To avoid discontinuities in the planning and cont.)process, each AERA system begins its planning process hefore anaircraft enters the control region. It is planned that AERA contro'sectors will be staffed by two controllers and that the airspacecontrolled would be several times the size of current day NAS serl.or-.

The productivity increases sought in AERA imply an increased trafficload for each controller, certainly a heavier load than an unaidedcontroller can handle, so the impact of a system failure on smoothtraffic flow and on safety will be much more severe than in the curronsystem. Thus, reliability of both hardware and software will hecritical; backup procedures will be more complex; and human farto:;.considerations pertaining to these backup procedures will he mnreimportant.

112

The basic philosophy for AERA system int-.grity i'; to des inn the AERAhardware and software so that faults of hardwate, software, andalgorithms have no effect on the performance of the automated system.Availability of this fault tolerant computing system at an en routecenter must be as high as state-of-the-art technology willermit--close to 100%. Our evaluation of computer technology of the980's indicates that a fault tolerant system in which a failure will

occur only under the most unusual circumstances can be built at areasonable cost by the end of the decade. This subject will bediscussed further in the Advanced Computer Program.

We have currently formulated a detailed AERA system operationa'concept and are in the process of developing a detailed functionaldescription to be included as part of the Computer ReplacementProgram. In addition, a real-time AERA test hed capable of exanininithe role of the controller in an automated environment, as well asdemonstrating concept feasibility, is being developed. A firstdemonstration of AERA's ability to accommodate fuel efficient Flightis scheduled for December of this year.

ALTERNATIVE SEPARATION CONCEPTS

Up until now, we have been discussing a rather sophisticated AirTraffic Control system dealing with well-equipped aircraft operatingin a controlled environment. Let us now discuss the other end -, thespectrum, namely, aircraft operating in the system with little or nointeraction with the ATC system. The Alternative Separation Conceptsprogram is investigating the feasibility of permitting properlycertified pilots to conduct flights in Instrument MeteorologicalConditions (IMC) on a relatively unconstrained, autonomous basis.

There has existed for some time the belief that in certain airspaceunder specific conditions, as yet undefined, that flight could heconducted in Instrument Meteorological Conditions (IMC) with thefreedom approaching that experienced under VFR. One such concept,conceived during the New E&D Initiatives effort and termed ElectronicFlight Rules (EFR), has been examined in some detail by LincolnLaboratory.

This concept relied on DABS coverage and the use of a collisionavoidance system such as BCAS or ATARS. The results of the Line,!.Laboratory effort indicate such an approach is technically feas,'!iand identifies the functional requirements that a "black box" wouIdhave to be provided to permit EFR flight in an en route environmnft.However, this study did not address the problem of terminal airspacenor did it consider procedural changes or airspace modifications whi,-hmiqht lead to a different set of functional requirements. Obvioisly,al ernative separation concepts are not limited to the EFR concept.,and further effort is required. A-, a first step toward cont inimlexamination of this issue, we will hp conducting an ASC Wnrk,;hp

the FAA Technical Center early next year and will involve allinterested elements of the aviation community. It is hoped thatnumber of promising alternatives will be identified during the coJrst,of the workshop and that the most promising alternative will beidentified and plans for development and feasibility demonstration canthen be formulated.

ADVANCED COMPUTER PROGRAM

You have just seen a sampling of some of the long term FAA programswhich are designed to accommodate the continuing growth in air trafficand the need to achieve higher levels of safety, performance andefficiency in the operation of the Air Traffic Control system. Thisneed, coupled with the lack of expansion capability of the current ATCcomputer complex, is the basis for the requirement for a significantupgrading of the ATC computer system by the end of this decade. TheAdvanced Computer Program is designed to satisfy this requirement andto lay the foundation for the automation system which will carry us tothe turn of the century, and even beyond.

The new en route computer system will be designed to provide thecornerstone for substantial air traffic control evolution in the post1990 era. The new computer system will:

o Provide increased automation of the planning and control functionscurrently performed by the air traffic control specialist. These"automated" decisionmaking functions will be closely tieJ to anintegrated flow management system that will tie together the enroute and terminal functions and the national flow controlcapability. All these functions will make extensive use of thediscrete address beacon system data link.

o Utilize improved weather data based on the introduction of lopplloweather radars in its automated planning functions.

0 Accommodate satellites for surveillance and communication, shouldthey become economically viable options in the future.

o Permit the use of cockpit display of traffic information to a11owthe pilot to assume additional responsibility in the cockpitshould this prove feasible.

0 Include the consolidation of terminal radar control into tho, onroute facilities should that be determined desirable.

0 Provide the capability to upgrade the terminal computer systemusing hardware and software modules developed for the en routeapplication.

WP wish to havp the nption of performing en routn and tPrminal

I 11J

functions in integrated faci !i ies or in i,) ate tac' Iit. ls t ', Icommon equipment and data ,ources. In vddition, thhis proqram willsatisfy the need for support fari lities utilized for devlopment,certification, and training.

The architecture of the replacement system will be such as to ensure .1functional reliability/availability approaching 100%. It will hecapable of being integrated into the existing air traffic contralsystem in a way which is initially transparent to controllers, p: ri;,and other users. It will possess sufficient flexibility to permit theair traffic control process to change to one in which the majorportion of the control functions are performed automatically bycomputers. It will accommoJate changes in computer technology durilqa long system life in such a way that no wholesale system replace-ment will be required in the future. It will accommodate anticipatelaviation growth and changes in the mix of aircraft using the airtraffic control system. Finally, the replacement system must ho ahl-to integrate new air traffic control processes and new air trafficcontrol data sources that may ar ise from thu evolut ion of Suppe r t in(navigation, communi- cation, surveil'ance and weather detection indforecasting systems.

The acquisition of the air traffic control computr replacement systemis being carried out as a major system acqusition as defined by D,ITOrder 4200.14A, which is the DOT order implementing the guidanceprovided in OMB Circular A-109. In accordance with that order, theprogram is structured as a four phase program with decision point; anlassociated TSARC's occurring at the start of each phase.

Competition is maintained throughout the initial phases of theprogram. The decision on whether to maintain two contractors throojghthe relatively expensive full-scale development, test and evaluat linphase will be made on the basis of a cost vs. risk analysis at theconclusion of the design and critical subsystem demonstration pha',..Present planning, however, is bein done on tho assumption that twocontractors will he carried throug the third phas,. Th:,, pprc.-assures minimum risk in achieving the desired syctem on the d ",schedule.

The initial phase of the program will involve a viripty ifprocurements, including technical support efforts and studies. T'-.activities may be categorized as: requirements study; long-termrequirements analysis and supporting technoloqy studies; manaqement;specification development for design concepts; ind su)liritat inn inr!evaluation of alternative computer system desiqn concepts.

A brief status report follows:

The requirements study team has completed a draft of the systemrequirements statement and that document i-, enterini) the cor-in , ion

i ':

process with a release of the final document scheduled fc)r Jnj i.y1981.

Technology tasks supporting the preparation of the specifiraticn anldesign evaluation criteria have been underway for some time andcertain key tasks have been completed.

The specification and evaluation criteria definition work was start Jlin March 1980. A draft specification is planned for early 1981.

The complete procurement package is scheduled to be ready to go oJt toprospective bidders in December 1981 with contract award planned forJune 1982.

Preliminary funding estimates have been developed based on pastexperience with similar programs, discussions with industry, ard ou."initial assessment of the impact of A-109 on the development andimplementation process.

These estimates indicate the cost for replacing the air trafficcontrol computer mainframes and software plus replacement of thedisplays is approximately $2.8 billion dollars including inflation.

HUMAN FACTORS

So far we have been discussing control systems, automation concepts,scenarios, advanced technology, and computers and their application tothe air traffic control systems of the future. But we have saidlittle or nothing about the other, and, perhaps the key element in thepresent and the future system, the human being. Human limitationsconstitute the major cause of accidents today. Systems of the f jturomust be based on proper allocation of functions between man andmachine. We have an extensive human factors program aimed at systemperformance enhancement and error reduction, both in the cockpit andon the ground.

First let us discuss the controller issue.

Controller Performance Enhancement and Error Reduction

Though controller related accidents have been rare, we havexperienced an increased number of system errors in recent years.An analysis of these errors showed that over ninety percent )fthem involved some element of human error such as inattention toduty, poor judgement, lack of coordination among controll,'rs,failure to properly identify aircraft, and poor communicatinrskills. These findings have resulted in the establishment ofcontroller performance improvement projects aimed at theelimination of error causes in the present system.

I l0

The introduction of advanced data technology into the AT systemnhas brought the potential of new sources of system error in termsof controller interaction with automation. Such issues as horedomand inappropriate intervention into automatic controls andinability to detect and intervene in automation failure situatinshave lead to the establishment of projcts aimed at definingappropriate controller roles, compatible with increasinqautomation. These projects in the aggregate are called theController Performance Enhancement and Error Reduction, or CPEERprogram. The goals of the program are to improve controlk rperformance and reduce error potential in the present and evovingsystem to improve system safety and efficiency.

The Present ATC System

In the present ATC system, the problems being investigated ar-e:

0 Improved information display that will assist thecontroller in heavy traffic situations.

0 Optimally designed sector suite to accommodate planned nearterm additions and training.

o Formulation and implementation of standard operatingpractices to help reduce system errors.

With respect to the evolving ATC system, we have a number of concernssuch as:

0 How do we assure controller confidence that automated

system elements are in fact doing the proper job?

o How do we assure that with increasing automation systeminduced errors and blurders are reduced?

0 How do we enhance controller performance in automationfailure situations and maintain vigilance during routinoautomation controlled operations?

How do we assure acceptance of the system by pilots,controllers and users as it evolves through increasinglevels of automation?

In this area, specific problems being addressed are:

o The role of the controller in an automated ATC environment.

0 The ability of the controller to perform his assigned jobin an automated ATC environment.

o The optimal design of the interface hetween tie olntrand the computer in an automated ATC environment.

0 The feasibility of the concept of "duto-controller"

analogous to "auto-pilot."

0 The impact of passive and active CDTI on the controller.

To successfully solve these defined problems, basic tools arerequired, and this represents the third area of the CPEER efforL. Inthis area entitled "Methodology and Facility Support" we have includedtasks such as:

0 Development and validation of controller performance andworkload measures.

0 Development and validation of objective and reliable AICsystem effectiveness measures.

0 Expansion of the current ATC simulation facility to emu! tt:system configurations planned for the future.

0 Develop the capability to provide full mission simi'i!i ,n

from the controller point of view.

Let us now turn to the pilot side of our human factors program ca'led

APEER - Aircrew Performance Enhancement and Error Reduction.

Historically, pilot error is involved as a factor in approximately 60%

of air carrier and 88% of general aviation fatal accidents. Recentdata from the NASA Aviation Safety Reporting System indicates thatpilot error is also a significant factor in aviation incidents.Further, as aircraft powerplants, airframes, and systems improvp, andas aircraft flight envelopes open to higher performance capabilitiesand lower minimums, the occurrence and consequences of pilot error areexpected to increase in proportion to other factors. In 1975, aspecial task forca formed by the DOT to study the FAA safety missionrecommended that the "FAA undertake a major safety research program toassure that future systems are designed around reasonable criteria forhuman error." Concurrently, the FAA Office of Systems EngineeringManagement had undertaken an in-house study to identify the humanfactors problems associated with both air carrier and general aviationaccidents and incidents. The results of this study were briefed toindustry and government groups and the feedback used to formulate aprogram of human factors research to supplement a number of ongoingFAA projects specifically aimed at reducing pilot error by identifyinqneeded but unsupported human factors activities. In 1978, theseongoing and new projects were pulled together under a central foc-3!point and entitled the Aircrew Performance Enhancement and Error

Reduction, or APEER, program.

14X

General Aviation Programs

In the general aviation problem category, we have identified and areworking on the following:

Problem: The high number of weather related accidents occurringbetween the basic private and instrument phase of general aviationpilot licensing.

Problem: Excessive pilot workload and inappropriate judgement aresuspected to be prime causes of pilot error in general aviationaccidents. The extreme diversity of existing and planned generalaviation cockpit equipment as well as the wide variation of pilotexperience and judgement capability will cause pilot error problems tnpersist.

Problem: There are no readily available general aviation human

factors laboratories in which data can be collected regarding generalaviation cockpit operation in current or advanced avionic equipment orATC system scenarios.

We are also categorizing general aviation accident data to determineadditional human factor problem areas. The leading problems relatingto workload or man-machine interface design will then he rank ordered.analyzed and potential solution areas identified.

Air Carrier Programs

In the Air Carrier cateoory we are examining the following problemareas:

Problem: Not enough is known about the potential benefits orliabilities of head-up displays in contributing to more flexible andsafer operations in air carrier aircraft during approach and landing.

Problem: The current generation of air transport aircraft has a verycompicated alert and warning system which could contribute to pilotjudgement error and may not indicate the correct priority of actionimmediately.

Problem: The introduction of advanced cockpit concepts and advancedACsystem improvements during the remainder of this century willimpose as yet undefined requirements on cockpit information processinqand displays.

Problem: The current status of cockpit standardization is unknown.The problems of non-standardization, vis a vis regulation, have n-trecently been assessed.

General Programs

The final category of the APEER program deals with general area;common to both GA and Air Carrier. In this category we are dealingwith the following:

Problem: Uncertainty exists as to whether mandatory altitude calloutsby controllers during Airport Surveillance Radar approaches wouldreduce landing accidents.

Problem: No fully acceptable, scientifically validated or widelyaccepted, systems approach to measuring pilot workload and performanceexists that will measure the utility of various cockpit configurations.

Problem: A number of accidents and incidents have been caused byaircraft taxiing onto active runways during takeoff, landing,or taxiing operations. FAA and NASA records indicate that 279 casesof this type have been reported over the past 10 years.

Problem: The rising cost of fuel has placed increased emphasis onmore efficient use of airspace. World-wide practices for thedetermination of separation standards have been under discussion inICAO. Although not yet fully quantified, it is evident that humanerror in navigation is a significant contributor to the failure of thoaircraft to adhere to designated route centerlines.

Problem: Now that the ability to provide traffic informaton to thecockpit exists, it is unclear what the pilots' ability is to use thisinformation or what the impact of his using it is on the ATC system.The benefits and liabilities of various types of traffic informationare unknown.

This last problem of the APEER program is the genesis for our programin Cockpit Display of Traffic Information (CDTI) which I will nowdiscuss.

Cockpit Display of Traffic Information

This is a joint NASA-FAA program with the NASA Ames Research Centerresponsible for basic human factors display and workload efforts, inithe NASA Langley Research Center responsible for air carrier cockpitsimulation. The Federal Aviation Administration Technical Center i-responsible for general aviation cockpit simulation using a low-cntweather radar presentation, and ATC system simulation and integrationwith the cockpit.

We are conducting this program in three phases which are aimed itevaluating the most practical potential CDTI applications as early a,possible. The three phases of the program are the Concept DevelopmpntPhase (Phase I), the Concept Validation Phase (Phase II), and the

Extended Applications Phase (Phase III). In early Phase I, whichconcentrates on the evaluation of a passive implementation, the pi>!twill employ CDTI in the role of monitoring appropriate surroundinqtraffic and assist with the non-passive tasks of merging and spacni.Phase II will validate the results developed in Phase I in anoperationally complete environment. The early potential application"from Phase I and II include:

o Controller assistance to the pilot in merging and spacini.

o Improved situational awareness.

o ATC back-up assistance in case of equipment failures.

o Blunder detection.

0 Airport capacity improvements in a high-density terminalarea with a traffic mix of diverse performance classes(Denver TCA).

0 Integration of CDTI and CAS functions through a shareddisplay.

1The latter part of Phase I and III will address extended applicationsof CDTI to additional aircraft classes and operational tasks includinqmore active pilot use of the displayed information. Some Phasp I andIII potential applications include:

o CDTI as a separation aid (not CAS) in low-density airsF ro.

0 Traffic awareness in a traffic mix of diverse performanceclasses (other terminal airspace than the Denver TCA).

0 A self-merging aid.

0 Advanced active functions (weather avoidance, 4D RNAV,runway occupancy, and other practical ATC tasks).

Although this program is designed to limit its scope to the CDTIconcept, there will be interactions between this program and other FAAand NASA programs. For example, the CDTI program will investigate thepotential interactions between a CDTI and cockpit Collision AvoidanceSystem displays, both as separate and integrated devices. Flighttesting will be accomplished using an experimental DABS data link as aconvenient pipeline to the experimental CDTI aircraft.

I51i

NONTECHNOLOGICAL ALTERNATIVES FORBALANCING AIRPORT/AIRSPACE SUPPLY

AND DEMAND

byHarvey B. Safeer

Director of Aviation Policy and Plans

I. INTRODUCTION

Aviation users continue to place increasing demands on the infrastructuredeveloped over the past decades to provide a safe and efficient airtransportation system. The growth has been most spectacular over thepast two decades, and while the role of growth is expected to be reducedin the future, nevertheless there will be growth. Capital investment infacilities and equipment and the introduction of major technologicalinnovations have enabled, for the most part, the infrastructure to growand keep pace with the growing demand. However, as the costs ofexpanding existing facilities and constructing new ones becomeincreasingly prohibitive, more attention has been paid to alternate, low

investment cost or noncapital-intensive techniques for accommodatingincreased demand.

These alternatives are generally of three types:

1. Alternative facilities to off-load congested airports(satellite, reliever airports);

2. Administrative (imposing maximum limits--quotas--on the numberand type of operations which may use a specific airport orrunway during a given time interval); and

3. Economic (charging variable landing fees, differentiated hy timeof day and by location; auctioning available landing and takeoffslots).

These last two measures do not physically expand capacity, but they canpostpone the need for physical expansion by promoting more intensive andmore economically efficient use of existing capacity.

II. SATELLITE/RELIEVER AIRPORTS--LOW CAPITAL ALTERNATIVES

The satellite airport program is aimed at reducing the mix of air carrierand general aviation aircraft in major metropolitan areas by makingalternative airports more attractive for general aviation use.

The basic theme of the program is that needed short-term development forcapacity and instrument training relief will be given high priority.Emphasis will be placed on projects that can be speedily implemented tn

-I

yield prompt results. Further, that priority will also be given o 'nrun development of the total general aviation and reliever airpo-t _yt.J-rin metropolitan areas and that such development is given specil oinphisi,in the Administration's proposal for extension of the Airport and A"rwlyDevelopment Act. Thus, the satellite airport program is a quick re)(sp n,,by the agency to provide improvements with follow-on efforts to !)rincluded as part of regular programming under the extended Act.

Facility installations would be funded through the FY-1980, FY-,: 111,iFY-1982 Facilities and Equipment (F&E) budgets. Other airportdevelopment projects would be funded with FY-1979 and FY-1980 Ai~polDevelopment Aid Program (ADAP) funds. Total combined F&F and ADAPare estimated at $93 million.

When completed, tne four year program will have included 86 sat lit,airports located near 56 metropolitan areas. Proposed projerts artestablishment of instrument landing systems, visual approach 1,indi,-systems, and construction or expansion of runways, taxiways and apro.'.Establishment of full-time control towers as a part of this program a,not recommended at this time. Shifts in aircraft operations will becarefully monitored, and tower service will be initiated where increlm c Vin activity warrant additional air traffic control. As part of thoprogram, however, temporary towers will be provided at two location,determine their impact with regard to attracting general aviationoperations away from the principal air carrier airports.

Identified for the initial F&E program phase are 24 locations forestablishment of an instrument landing system. Equipment will heobtained by reprogramming vacuum tube ILS replacement proJPcts !iudj d!in prior years. The proposed FY-1980 budget contains $11 million fzraccomplishment of thp deferred ILS replacements. In FY-1981/19?,,$23 million funding level is proposed to complete tho F&F portion o'program. For ADAP funding, $31 million in FY-1979 is recommernd ,projects at 49 airports. In FY-1980, a $28 million requirement i-recommended.

Major airport development projects include the allocation of $11 in! Ifor public acquisition of four private airports and $V million for 1,11!purchase and initial construction of two new satellite airports atReserve, Louisiana (to serve the New Orleans and Baton Rouge ar,2 in Iat Baytown, lexas (near Houston).

Program details are provided in Appendix A.

III. ALTERNATIVES TO CAPITAL INVESTMENT

The alternatives to capital investment can be grouped into thr,categories.

Administrative Measures: Access to c:ongested airports ', ,*,,, .

through administrative fiat. The restrictions may be selectively af)p. ois,to specific categories of aviation, to certain periods of the day, and t,some but not all runways of an airport. The establishment of a "q&,ta"system that limits the number of operations per hour is another form ofadministrative control. Airport curfews (usually imposed to deal withthe nighttime noise problem) are an extreme form of quota.

Economic Measures: Airport congestion is controlled through a pri, 11isystem which imposes higher charges during periods of high demand t .inother times. This category can be further subdivided into measui- ti.'tplace a peak-hour surcharge on aircraft movements and measures thot i'vthe surcharge directly on air passengers.

Hybrid Measures: Access is limited hy a combination of adfninA'iand pricing alternatives. A quota system limiting hourly operati,)n,could be combined with a peak-hour surcharge or slot auction t) athat the available time slots are allocated efficiently.

A. ADMINISTRATIVE MEASURES

The first-order and short-term effects of administrative altera ..are straiahtforward and predictable. As the number of flight; it i:airport is reduced by imposing a quota on the number of flights,

scheduled or by banning specific types of operations, congection ivthat airport decreases. Because the relationship between ,iirpor!demand and airport delay is very nonlinear, a carefully chisen "'T,,on operations at a severely congested airport may drastically rede,delays without a significant reduction in the number of flights .Therefore, quotas and other administrativw measures have boen Iancontinue to be) particularly attractive as a means of deal in ,v.,and effectively with airside congestion.*

*In 1969, the FAA in the United States imposed hourly quotas -,nscheduling of operations at the three New York 'ity airports, '..'.,International in Chicago, and Washington National. The gut ,, ibeen generally credited for strongly ameliorating the traff i,congestion situation at these airports. Developments sinco !9%l i,;

made it possible to eliminate the quotas at the J. F. Kennedy ,'Newark Airports in New York. However, the System continu,effect at the other three airpo-'ts.

In the long term, however, the impact and henefits of purelyadministrative measures are less clear becausp they offer noassurance that economic considerations will play a role indetermining who will use a demonstrably (by virtue of it beingcongested) valuable facility or how this facility will he developelin the future.**

Once a user, for one reason or another, has heen denied access !It) L;airport, he has no way to prove that any given time slot is morevaluable to him than to its present occupant. As long as the pre~eonoccupant is willing to pay the fixed landing fee that the airportcharges for the slot in question, everyone else is excluded. Thereis no opportunity to "bid up" the price of the time slot to reflertits value. Even where the time slots are periodically reneqotiat,,d,there is no way--in the absence of a pricing mechanism--ofascertaining that the right to land at any specific time will f,.those who most covet that right. In fact, it is not clear thatreassignment contributes anything to increasing the economicefficiency of the time slot allocations.

Administrative limitations on airport use, by keeping demand wilhacceptable bounds, may assure the relatively smooth operation ct thffacility and the lack of severe congestion into the foreseeablefuture. But with access to the airport restricted and with poteWusers unable to indicate the true value to them of future airpn-texpansion, a false signal is conveyed to the government and thepublic alike. In effect, by arbitrarily constraining demand,artificial equilibrium conditions are created which, in the lit .-are likely to distort the nature, quality and cost of thetransportation service provided.

In summary, purely administrative me3sures, while effective an'probably desirable in dealing with short-term congestion probleir,tend to be strongly biased toward maintenance of the status qun whenused over a protracted period of time. Because economic value i,; nftfully considered in allocating time slots, current users cannot hodisplaced by others who may derive a higher economic value from II,.same time slots and the airport cannot obtain through economicmechanisms the information required to determine the need (o,- i,,thereof) for capacity expansion or for an improved (or, for thuimatter, a reduced) quality of service.

**The purely administrative case is one in which right,; for 11wthe runways are offered and time slots are allocated either byexecutive fiat or through negotiations among users. In eithersituation, it is assumed that no explicit or implicit econenicbidding for landing rights and time slots takes place.

B. PEAK HOUR PRICING

The use of economic incentives rather than administrative cont',scould alleviate the long-term allocation and development problems ilthose incentives could be tied to the true costs and benefits ofaccess to the airport. However, this is not a simple task hecii.s,there are both private and social costs involved.

Once the level of operations reaches the capacity of the airpo,. '!waddition of one more operation slows the flow of all operation',generating a delay for every aircraft attempting to use thefacility. This additional marginal operator considers only hiK .-'ndelay in deciding whether or not to use the facility and faili toconsider the impact on all other users. Thus the true socialmarginal cost deviates from and is higher than the private rarrjni Icost.

C. HYBRID MEASURES

Hybrid measures use a combination of administrative and econoIrTii,techniques to control demand. For example, the operational surch~iujon general aviation movements during peak periods which was imposedby the Port Authority of New York and New Jersey in 1968 coupled withthe "quota system" that the FAA imposed in 1969 created such a hyhr;,Ienvironment in the New York area. A similar example is thecombination of economic charges imposed by the British AirportsAuthority and the quotas imposed by the United Kingdom's 'ivilAviation Authority at Heathrow Airport.

D. PROBLEMS IN APPLYING THE THEORY

The basic arguments in favor of a time-varying structure of 3i -po0fees and some general policy guidelines (for example, fees shoproportional to the magnitude of the marginal delay costs) forsetting these fees are well understood. Current pricing pIi,congested airports are far from optimal from the standpoint of

economic efficiency.

The next natural step should be to establish price structurcs !,,ritspecify airport usage fees by type of operations and by timf, of ,lv.Such fees would maximize the net benefits that society derives fromexisting airport facilities. Unfortunately, there are a number offactors that immensely complicate the determination of such a p.i,-structure, ranging from analytical difficulties to question, fpolicy toward specific segments of aviation. (There may he va l1policy reasons not to seek economic efficiency as a primary ,ni

1. Determination of Equilibrium Prices

Although models exist to estimate the marginal delay costsassociated with any given flight demand and capacity prof ;I,.there is no guarantee (in fact, it is highly unlikely) thatsetting congestion tolls equal to the marginal delay c )s!, underany given status quo will lead to equilibrium conditions. ff ahigh fee equal to the current marginal delay costs were impss-?on operations conducted at peak traffic hours at a specif:rairport, at least some of the peak hour operations would rnluvcoff-peak hours to avoid the surcharge. This would lowermarginal delay costs at peak demand periods, and converselyraise marginal delay costs at off-peak periods. Runway jsau,fees would then have to be readjusted (lowered at peak hoijrs in"increased at off-peak hours) thus luring back some of theoperations that were previously driven away. This sequento,could continue ad infinitum unless some means of computingequilibrium prices exist.

In short, there is no information on the sensitivity of airpor!demand to changes in runmway usage charges. And without -h-TVinformation, the effects of any specific price structure on t'Wpattern of demand for airport operations cannot be predict(.d.None of the mathematical models reviewed considers demandelasticity, even at the theoretical level. They assume unf,',demand which ignores the relationship between the number ,operations at peak and off-peak periods and the respectiv,levels of runway charges during these periods. This is on(the most fundamental deficiencies of the analytical work tt rexists to date.

2. Network Effects

The air transportation system requires a complex netwo ifacilities which are often interdependent. Congestion a !, yone airport cannot be considered in isolation but must !(analyzed with due consideration to congestion at al, othrairports with which this airport is linked. It is conceivi <',that pricing systems developed separately for two interconn,..airports would cancel out each other's intended benefirialeffects. Ideally, therefore, airport price structures shidetermined for networks of airports rather than for each iihairport in isolation, which requires a degree of cnnple itIvbeyond current analytical models and techniquos.

On the other hand, one can argue that no major airport v,,: iv-,an overwhelming fraction of its flights from any sinql,,and, consequently, that the interdependence is loose ,,n,;,permit examining each air-port separately. Neither pwJI' 1,, i,be proved or disproved at the present time.

However, the interconnecl ion of airport cannot he 1i1mi-,,o, ,easily from an individual airline's poinL of view. Bec-ui;-, ofthe need for high aircraft utilization, existing airlinescheduling patterns are relatively insensitive to isolatedchanges in landing fees. Changing the arrival or departur,, tiim,for a given flight at any particular airport would mean chanqiniflight times at all other airports that the aircraft servP,_--no!to mention the "ripple effects" on all the connecting or"feeder" flights by other aircraft which may have been set. tnconform with the first aircraft's schedule.

However, this is consistent with the concept of marginalpricing. The example implicitly states that an airline miv [,

willing to pay a high fee for the privilege of having itsaircraft land and take off at a particular time at a specificairport. Another airline or a general aviation user, who doesnot place such a high value on the time of a flight to th, ,-am,airport would not pay the high fee but would use a differeonfacility or fly at a different time. The result would be ,ho;congestion and improved utilization at the airport in qu,,.; ei.

However, the main point cannot be ignored: to predict anairline's response to a time-varying price structure at iparticular airport, the repercussions of each possible ,-hngo .the complete schedule of the airline must be considered. -t i.,unlikely that an airport planner or airport economist 4an kthese estimates. Therefore, knowledge about an airline'selasticity to runway fee changes will probably be uncerta,,a long time to come. And, as a result, the impact of pr 'structures on an airline's behavior can be determined 'napproximate manner rather than be estimated from specifi :V iinformation.

3. Recovery of Facility Costs

The analysis of time-varying runway fees that would mamin.'social benefits did not place any minimum acceptable limiL, .ethe total revenues that airport authorities should collert frnsuch fees. Yet such minimum acceptable limits do exist inpractice because the airport must recover at least some of 'f

maintenance, operation and construction costs. Although tn,degree of expected recovery varies widely not only from ono,country to another but also from location to location withi:,icountries (especially in the United Stat's), there i-, , itpolicy in developed countries that major commercial airport',must be more or less self-supporting economically. Thi-, hI',several consequences for pricing structures.First, the miniI,tamount is usually substantial in absolute terms. It is,therefore, entirely possible that the airport's basic ne,, c-,;

exceed the amount col lected under a marginal delay cost pt Fl.policy designed to optimize the social utility of the airpo-t.Optimality may in fact be precluded because the landing f(,eneeded for financial support are too high and eliminate flightthat would have taken place if charged only for marginal ,lliycosts.

Second, it is very difficult to determine what the long-t-wn

costs are that must be attributed to and recovered from ai.-J!,;users. For instance, if onc of the airport's goals is toaccumulate resources for future expansion and improvement,, ifmust know what such future changes will be and what they willcost. However, in most cases this information is unknownbecause future expansion plans are contingent on future aKi;'prdemand (which in turn is affected by the very set of usage fe(,ethat have to be determined). In the United States thisdifficulty is bypassed, as a rule, by requiring recovery aft-,the fact; after a facility has been built, its users pay fe-that amortize-F-the cost of the facility over the estimateJ ..')inof its useful lifetime.

Third, the proper way of allocating facility costs among u.'-'is very controversial. Presumably, if marginal cost pric q; ,o

to be the norm, each user must pay for the additional wea- inv itear" (marginal short-term costs) and the additionalconstruction costs (marginal long-term costs) caused by hi,of the airport. But, of course, it is practically impos;fi,,really determine what these marginal costs are. The presorlsystem of computing landing fees, primarily based on airraftweight, can be viewed as an attempt to deal in a practical N',

with the problem of charginq for marginal facility css.

There is, however, another inteprotat : )n of weiqht-h v,-, Iwhich, although not central to the Aiscus;.ion on thi., )r ,is directly related to airport cnnqestion. Airside facili' i-.have very high initial (i.e., construction) rests, hiA on,- 'horunways and taxiways are constructed, the costs rausd hy inadditional operation at the airport are lose to zern. A-,

as the facility is underutilized, it is ti the .irport,advantage to invite additional users to tho la i, , - , ',revenues. They do this through a disrilnpna'rv pri-e ', i,based on willingness or ability to pay. In th's ,- rtsx.aircraft weight is a proxy for willinqn, ,r .i.i " I I y t r)light, general aviation aircraft whos,, np rit'), , I n Ipresumably afford a large -unwiy foo T,,, r -t-. w) ,much less than a commercial "t.

While this pr icinq policy miy be ippropr 1t, fo)t jni f1' ,and uncongested airports, it i, -ounterprodr! iv, for , o w;facilities. By pricing the fari ity aru o (I in to i. ii°.,1 !-

weight, it effectively encourage"s those users who are l,-swilling to pay full costs and who, at least on the average,contribute less "transportation value" per operation to soc ietas a whole. In fact, the only deterrent to airport use jilde.present weight-based systems is the prospect of delay it acongested airport. Thus, a user who is able to tolerate his )jr,delay will use the airport without considering the delay ,.)Osthat he imposes on others.

4. Reluctance to Change

Pricing based on marginal delay costs, in its pure form, ctar-i,'each airport user with the costs that user imposes on others.An airplane landing at a busy commercial airport during itraffic hour could be charged $1,000 or more, depending on t' =

level of congestion and the mix of traffic at the airport.

Such a pricing system clearly does not consider the "abil 1y 1.pay" of the user. Whereas a $1,000 charge may be a smallpercentage of the total revenues of an intercontinental fliqtof a B-747 Jet, it would exceed the total revenue of mostcommuter carrier flights. In other words, marginal delay costpricing could result in a schedule of charge whirh may be harl(or impossible) for certain kinds of aviation or types offlights to pay. General aviation flights would feel the iqnn-of marginal delay cost pricing most severely. Flights ofregional and "third level" (commuter) carriers and, to . .extent, short-haul flights by trunk carriers would also bostrongly affected. In short, marginal cost pricing wnull :likely eliminate flights for which airport fee, represent isizable percentage of the value of the flight to the ai-rn f'operator.

Although this is the purpose of marginal cost pricing--f.eliminate those to whom the value of using the airport i-than the costs they impose on others--the imposition of oee,without consideration of ability to pay is such a drasti,-departure from prevailing practices that it may appear t,)discriminatory and unfair. Those users who have relied on Vi,traditional low cost policies and who have developed vstfeinterests in their continuation are the major impediment t. t!a ,adoption of marginal cost pricing at congested airports.

Likewise, airport administrators can hardly be expf<c&ed t o .'Iin one single step pricing policies which are so differont f,-i,the ones they have been accustomed to and with which they iiv,had long experience. Their reluctance in this respect ojl I hoincreased by apprehension regarding the exact nature of ;h;-tand long-term reactions to the new policies by the variousaviation interest and pressure groups. In addit1TrnlT-

---gop_ nadi-o -T

inability of analysts and of economic theory to determine inadvance equilibrium pricing schedules and to forecast theprecise effects of margina' cost pricing on airport uagE raistessome financial risks that could result from a departure frompresent practices.

To evaluate the potential economic (and political) risks and r],possible benefits of peak hour pricing, actual experience ar,not theory is needed.

E. AUCTIONS, EXCHANGE MARKETS AND THEIR APPLICATION

Auctions, in one form or another, have a long history dating hbk ancient times. They gained prominence in the commodity exchanges .the sixteenth century, and are still found in active use today forart sales, some commodity markets, oil lease sales, the sale oftimber and mineral rights by the Federal Government, and the sal e2Treasury notes. Auctions are very familiar and easy to descrihe, htdifficult to define precisely. Perhaps the most notable feature 0'auctions is the passiveness of the sellers. While buyers areactively bidding for the item(s) that are for sale, the seller wa,*passively for the highest bid, or accepts the first bid in a Dutchauction where descending prices are announced. In most auctions, tlKsale is final as soon as a winner has been determined. This is agreat timesaver, and perhaps the major reason for the popularity nfauctions. In other types of markets, buyers and sellers engage inprotracted negotiations; they may break off negotiations withojtreaching an agreement or contract; they may seek a better deal bytrying to find other traders who will offer more (accept less), thnvmay seek to resell items just purchased; and so on. An auctionusually provides a speedy contract between a seller with an item tosell and a buyer interested in that item by allowing a large numherof interested buyers to bid simultaneously for the same contract.The auction may be of the sealed-bid variety, in which case thebuyers are unaware of the buyer's bids; or it may be "open," in whicrcase all the buyers know each others' bids. In either case, once theauction is closed, the buyers must live with their mistakes. Theymay have paid "too much"--that is, they could have obtained the iter'for a lesser price. They may have failed to obtain an item by notbidding enough. In either case the auction does not provide forrecontracting. Buyers can attempt to negotiate privately with othebuyers in an aftermarket if one exists, but as far as the originalseller is concerned, the auction contract is final.

1. The Application of Auction Procedures to the Runway SlotATlocation Problem -- - -

The justification for rationing slots by price lies in thospirit of deregulation currently being espoused by the CAB.Through differential ticket pricing (in which the airlines have

inability of analysts and of economic thpory to determine inadvance equilibrium pricing schedules and to forecast theprecise effects of marginal cost pricing on airport usagE 'aise ,,some financial risks that could result from a departure frompresent practices.

To evaluate the potential economic (and political) risks ani '. opossible benefits of peak hour pricing, actual experience aFr,-not theory is needed.

E. AUCTIONS, EXCHANGE MARKETS AND THEIR APPLICATION

Auctions, in one form or another, have a long history dating h.3ck toancient times. They gained prominence in the commodity exchanges *,the sixteenth century, and are still found in active use today forart sales, some commodity markets, oil lease sales, the sale oftimber and mineral rights by the Federal Government, and the sal eTreasury notes. Auctions are very familiar and easy to descrihe, bdifficult to define precisely. Perhaps the most notable feature 0'auctions is the passiveness of the sellers. While buyers areactively bidding for the item(s) that are for sale, the seller wai*passively for the highest bid, or accepts the first bid in a Dutchauction where descending prices are announced. In most auction,. thsale is final as soon as a winner has been determined. This is agreat timesaver, and perhaps the major reason for the popularity nfauctions. In other types of markets, buyers and sellers engage inprotracted negotiations; they may break off negotiations withoutreaching an agreement or contract; they may seek a better deal bytrying to find other traders who will offer more (accept less); thevmay seek to resell items just purchased; and so on. An auctionusually provides a speedy contract between a seller with an item tosell and a buyer interested in that item by allowing a large numberof interested buyers to bid simultaneously for the same contract.The auction may be of the sealed-bid variety, in which case thebuyers are unaware of the buyer's bids; or it may be "open," in whichcase all the buyers know each others' bids. In either case, once theauction is closed, the buyers must live with their mistakes. Theymay have paid "too much"--that is, they could have obtained the iter'for a lesser price. They may have failed to obtain an item by notbidding enough. In either case the auction does not provide forrecontracting. Buyers can attempt to negotiate privately with otherbuyers in an aftermarket if one exists, but as far as the originalseller is concerned, the auction contract is final.

I. The Application of Auction Procedures to the Runway SlotAllocation Problem

The justification for rationing slots by price lies in thespirit of deregulation currently being espoused by the CAB.Through differential ticket pricing (in which the airlines have

only recently been able to freely Pngaqe, the cosO of rhvarious slots can be passed through, at least in part, t.) t'air travelers. Consequently, demand for travel at va~ioJs time.of day will affect, and be affected by, ticket prices; intheory, at equilibrium, an economically efficient allocation )ftravel facilities is a likely result.

Two types of auction procedures have received most attention.

Under one, the airlines bid directly for slots or bundles ofslots; under the other, they bid for the -ight to select a slotor bundle. Each of these procedures can be implemented within ,variety of auction mechanisms.

There are also several fundamentally different ways of dea'in

with the results of an auction. These results can be taker, 3,final, obligating the airlines to certain activities over , ",

period of time. Alternatively, if the results of the auc; , 'involve minor inefficiencies which the airlines wish to rect;*y.they can be allowed to revise the auction outcome throughprivate aftermarket (which deals in "property rights" for

slots), or to "back out" of certain tentative commitments 'r!-

which a secondary auction may be held), or to revise th,-i-

original bids (which consequently revises the rp-sultinqallocation of slots).

A principal feature of the slot allocation problem is the fa t thi p.',iions slots are used in pairs (for every landing, there is an eventj]:takeoff). Therefore, for example, any auction procedure under which a-airline might find itself allocated an odd number of slots should b,viewed as the initial part of an allocation mechanism which permit,after-auction adjustment of the slot assignments.

There are a number of attractive aspects to procedures whir,involve the airlines in bidding for the right to select slotsrather than bidding directly for the slots. Indeed, if all

operations slots were distinguishable (that is, were attached to

specific times), such procedures would probably be the bestavailable. However, there are two critical disadvantages tothese procedures. They are informationally complex, calling for

each airline to make a large number of individual decisions inan uncertain environment. They also are highly sensitive timinor misperceptions of demand. As a result, a slot (moreprecisely, the right to choose this slot) in a given one-hourperiod may sell early in the auction for an amount substantiallyhigher or lower than the price of an identical slot later in the

auction-after the total demand for such slots becomes clearer.Hence, the variance in price of identical slots may beunacceptably high.

Alternative procedures involve biddng fir'ectly for slots. Most

such procedures can be viewed a,, dynamic-pricing mechanisms.All the slots within any given one-hour interval (of course,intervals other than one hour can also be used) are allocatedsimultaneously. Coordination of allocations across intervals,in order to satisfy slot-pairing requirements, is handledthrough subsequent adjustments to the initial within-intervl'allocations.

The allocations of slots within an interval can be determined ydirect adjustment of prices. Assume that the expressed demandfor slots exceeds supply at a per-slot price of zero. Then theprice can be gradually rais-d; demand will decrease as the pricoincreases. The "equilibrium" price, at which demand firstequals supply, can be tentaively established, and the slot -

allocated accordingly. An eltivalent version of theprice-adjustment proces ha-I each competitor submit aprice-quantity demand curve, or (under an appropriate assump' ionof convexity), a list of bids for an initial slot and foradditional slots. The equil hrijur price will be equal to thehighest rejected hid.

A drawback to this approach can he seen tnrough an example.Assume there are bids of 8, 8, 8, 7 and 6 for a supply of fourslots; further assume that the first three bids are all entert.i

by the same airline, and the fourth and fifth bids by twe hairlines. The first airline will then receive a bundle of ,-',ot-of (subjective) value 24, it a cost of 1, units. However, iralternative strategy for this airline would be to misrepresentits demand, entering only two positive bids of 8 units each.Slots would now he priced as free goods (there would be noexcess expressed demand), and the airline under considerat ,nwould net a gain of 16 units for the two slots purchased ratherthan only 6 for-he three slots purchased.

An alternative pricing merhanw'm involves eliciting bid lists,awarding slots accordinq to thr -,: hest bids, and charoing eachairline the sum of the hids 'other than 'to own) it has causedto be rejected. An airlin, w~th k winninq bids would be charg:'the sum of the k highe t r-,lecti. Lids" obviously, this cannotexceed the total amnrt bW tv the successful bidder. In thFepreceding example nhiechar<,,i: wonild charge the same amount,6 units, to both winnirg ' ire. The ;n(entive formisrepresentat ion nf Ipman! ho . , P iinated.

An open adjustment peri,r!, -r wt irh ,;,- irlin are allowed :,r

revise their bi,I lists in view u f th. avr,,-Intrrva1 slntallocations, could be permt.d I-) far 'lit it, rl-,olution nr thoslot-pairing problem.

2. DOT Auction Proposal fo a;njNn at '_(01' Airplort

The DOT is proposing a ,iwi riiii ai1 !u ionpredu'Nat ional1 A irport w it h orpaira to irt i on' f or 1 1r c r rcommuter reservations,. lrvtion itic inns we id a 1, lwcompetitive pr icing of the roeor vat. ion-. For hour,, whonreservations are in h iq~h demcir d t h prir e o f a re '; r va,would be high. Converse 1, fcx hour'; whr, there( i,, I,-,- io ;Ifor reservat ions , the pr if-owa vvid drop I award z.'ro arrd war,1become $1 .00 if there is; (lxces,, s upply. T he ;spe c ifi - p e-, .proposed by DOT consists of Noa oloererr ts: a ' reservationauction" followed by a continuously opcri'eer vatierexchange." The auction pr'OCNdUrn termeod thoe "lrailinq Pni'.Auction" is described in considerable detail in i -oportentitled "The Allocation of Runway Slots h ''~n prepa'>'i :VEcon, Inc. The prop' osal For air carr ier: is i-, F-)' IwsJsimilar procedure would be us~ed for air I axte',.

a. Each air carr ier want inrg an TFR rosorvat ion w,-mi hoerequired to submit, by Janujary 1 an.1 July, ' of eaich yiesealed bids f r the ros'orvat ion, that it de, rt- fn, ions ix month per iod beq inn inq the fa I 1,%qi rij Arpr i 1 0r t OW.For any gi vein hour, an air carrier ,oiild hid1 for ' n'more reservations;, a; at the s.ame pr-ice or it atm0 dprices. Each hid would he reqiji red to ident ify tiheand spec if ic hour for, which the reservat ion is reque;1 ed.A carrier may submit bids, for as many recervat 'on' a'desired in the same or d iffereiiI. hojr- ,

b. Within 7? hourQ after the, rcce p1 dead] no fnr in i

bids, the nOT wooid make, pub] i c to iocirobatp demln,for each hour (but riot the indiv io! i a H ine rpinnficurves), the market pe ce for Pach hour 's -, lot,;, anJconditional al locat inn of the number of slot.-. that thc'.would have obtained is a result- of their hids. The,aggregate denrand wouldI h- a-rived (it 1by idig toqeot re,-each possiblo price, the i jmber of slots that ear h ')k-desi res . Fi t her tho tot al demand would he equa 'a I o .o't

than the supply of s lots t h it are 3va i 1 able, in whi ich ,hallI demands, ire met at a dol lar pr icor the temnand w,-u 11.'exceed the avail ate suipply. in the, caise of o xces', lar maiolthe market pr ice word 11 ho establIi shed at. the hhetprice that is cel~cte! F or example, at. Wash inqliranNational , thc. 3'*'- cr~ ',-. qjurt ifor any houir woruldinit ial ly he seot it 36) -)lot,-. 'h-' h i qh--t 1)idde1rs -,:,, ' i36 sIot s would, obta i n t i-r a t t he h i hes pc l - )id hit

i nsuf f icien t to oht i in thrn 36th s1 -1 ,t wh ic h wouuit h tr3 7th h i ghest pr icr- h i d. Tner'> nror' l Ircc's' riwould pay the, same price ror eaich rinna i!o 'hil 1-

c. If the condit fona3 a! io im ind pr ' ' aye t Al" ' m,a l bidder s, an eff ien L, coinpet iti vi' oqu i br 111 -fI lhave been found. n h' e 1 1 1,s' O. :awarded bas ed upon 'ie ii al bid pr ic. ( w 1expected, however, thait. lill', of thy rvdr, n i ntsatisfied with thie (i' ~i

d. An air carrier wishiriq to submnit new bius wrwi, i, 1 I)j,to do so with in 12 hours after the hMT) annouji(,cs tentf I1 vawards in accordance with proposed Sect ion '.13-() l3'Only air carrijer-, who hive made a hid in thc i-A tjn !otbi dding would he al lowed to submit a bid in ihe~oround. During this additional round of uiddinq, awould hid on any number of reserv3t mnn,. in .ory hoijr. Thosealed, secret, ond~ ido 31 bids w ilil1( ht, 3c~nr ions ame procedure rmpl 1 iyod ri t he f i rs t rjutI, is r rn,cond it ional1 all1oc -it ions, 1.ra di ng pr *:t~-; )n o t ta Ira.would be announced. This process is repeater! tha!instead of a -n-time auction, tht-rf wounld !bf a le eauctions. Each iterat Ion wou'd increas e t he of'Irmna' ionavailable to the bidders, each would add to ti iddinsight into the demand pressure a)Ver a' 1 hnu" Of teday. If , at any step, no bidder wi,,hesl to rhiijnqe V iJ,then the process would termi nate at. ,r) eo i ihrom ii st 01.,

e. After any 72 hour per iOd dur-ing which no il areceived by the DOT or if the only bi+', r .co1 1, w- yin72 hours since The DOT l1ast made puthl it the nu,:6'o o,reservations requested, are minimall y 1,!ter-n! frow Ilast !)ids rceci Verl b '1ding would -. lfs' and pr ic- A),

est ab 1 is bed . A 'o o nium di fferenc. i say tr in n '< ilr?the number of hi ds ~bni e.The DOT propwoo J

minimum bid diffrrenc(' ,)rocedure to estabi ish t"equili~br ium" that wi I, pre':lude n inor flu~ctu,1 Jonj

carrier lbids unnecessirily extendinq the t ime; taccomplish the s lit aiust ion. Base d on the FAA Xpo io'with the auc-tion cirnla~l ion performed ls e iy hDOT proposes that a irinirnum difference he otalse20 percent or less. For example, if tht! lacst IOprirmoIribid price was $40 per reservation and the mlnI,',submitted bid for that hour was $47, the new 5id wm': h.,rejected and biddinq would be cloi-1 s) inre tho Iiffr('between the two hid prices ($17 - $40 ewhich il $)than 20 percent of the last bid pr ir'> A mi r nia' ly1different ctibmissien of bids wot, ho ai chanlo in20 percent or lesof the ovorall a wa - +1; For oxafnp', ifonly 20 bids are sihmi tteod (out. of a I ota' 1)? bids,avai labia in any day for resi'krviat inns, ti iddinq wnu', n fsince that is be 1ow ?)percent if t his' t i t a n hf nrnbu'r,avai lable. In t' r 1 ancos t ho'i Id- wo)ill frequi ePd to) v~c' p! t, JW,', iwoD' n i

n ti iqa t ; )ns- o f ; i r I it in '' I lal, ,

f. Alternatively, the au. tion would ho ,,erininato:i awt l'recent conditional award-, and prices finalized, ifparticipating bidders inform the DOT within 'I ,i, ur, ifthe announcement of cond it ion al awards that thy , ! tterminate the auction.

g. The DOT would notify each successful bidder in wr i 6i,! ofthe amount of its accepted hid prices. The hid ler wonud h-required to submit to the DOT payment for one-sixth (f thtotal amount accepted within 14 days after it ha- rr.-,ive ,such notification. The balance would be payabl, in f iv,'equal increments starting on the first day of thereservation period. Succeeding payments would be nealo it

the first day of each month until the balance i5 nai

h. At any time between award of reservations and the oen iC

the time period over which tK-y were purchasoJ, re,"v tieiowners could sell their reser,,ation to other use rs thrijlia DOT reservation exchange. Any reservat i,)n ow vner q i',,to sell a reservation and any air carrier or air taxiwishing to purchase a reservation would notify thI, DOTwriting of its intention. The DOT would maintain a Iavailable reservations and desired reservations and S1,prices, but the names of the bidders and offerori, woIo i -in'be disclosed until the transaction is completed. Th i,,would eliminate special deals, collusion or rcon',piracyamong the members of any class of user. The DOT would,available reservations open to bids for" a minimum of24 hours after notification of their availability. The IOTwould sell the reservation to the highest hiddtr; hiwevr -,

the seller would not receive more for the particilar ,.1than it paid. The excess amount over the oriqinal p "'laprice would be paid to the Treasury (unless authorityobtained to place the money elsewhere, as previouslydiscussed). The seller would be absolved )f any payne;-trequirement associated with the relinquished reservat ionprovided that the resale price received by the DOT is "qqm'to or exceeds the seller's payment obligation. Air caririorreservations could be purchased by air carrier , rscheduled air taxis while scheduled air taxi res "vat io,,could be purchased by scheduled air taxis only.

B I B LIOGR A PH Y

1. Park, Rolla E., "Congestion Tolls for Commrerci-i' hp r

Econometrica, Vol. 39, No. 5 (September 1971). pp. C~~A

2. Fitzgerald, E. V. K., and Anouryn-Evans, G. R.," 111ho 1coom~k-Airport Development and Control," Journal of Transport F-onnmi-c,, 11oPolicy, Vol. 6 (September 1973), p7~~

3. de Neufville, Richard and Mira, L. J., "Optimal Pricinq V.Air Transport Networks," Transportation Research, Vol. 9A,pp. 181-192. -

4. Douglas, George W. and Miller, J. C. III., Economic PRogula, onDomestic Air Transport, The Brookings Ins ti j EL CT6, -W,1c7{Cn-qt.Y

5. Carlin, Alan and Park, R. E., "Marginal Cost Pricing -if A ;por'Runway Capacity," The American Economic Review, Vol. 61 :.i7O).pp. 310-318.

6. Yance, Joseph V., "Movement Time as a Cost in Airport ()po''iIJournal of Transport Economic Policy, Vol. 3 (January i?pp. 28-36-.

7. Anon, Airport Capacity Handbook (Second lEdit ion), pr'epar'~t>- Y

Federal Aviation Administrati-0n by Airborno Instrumont I '11)Farmingdale, New York (1967).

8. Hengsbach, G. and Odoni, A. R., Time Dependent Estimates ni Ple,~and Del ay Costs at Major Ai rportsM..T. F7--iT Tranporta_(TrnLaboratory Report 75-4,T_mr-1dge, Massachusetts (January '.97:)

9. Koopman, B. 0., "Air Terminal Queues Under Time-DependentConditions," Operating Research, Vol. 20, No. 6 (Nnv,2mhrJ")ci1W!1972), pp. 1089-l114.

10. Eckert, Ross D., Airports and Congestion: AProblem of MiplacrdSubsidies, American Enterprise Institute rPulrlr7~'rWashington, D.C. (1972).

11. Levine, M. E., "Landing Fees and the Airport Congestion Proi i-'rn.Journal of Law and Economics, Vol. 12 (April 1969).

12. Little, I.M.D., and McLeod, K. M. "The New Pricing PolIryBritish Airports Authority," Journal nf Transpoirt Ecoinomic .ro

Policy, Vol. 5 (May 1972), pp71-1

13. Wilson, R. , "A Bidding Model of Perfect Competition," The kev ww (JfEconomic Studies, Vol. XLTV(3), October 1977.

14. Smith, V. L., "Bidding and Auctioning Institution,;: Exper inent iResults," Bidding and Auctioning for Procurement and Allocation.Y. Amihud (d)NeYork NewY o r k ni eT7_yTrj7_ 176

15. Plott, C. and Smith, V., An Experimental Examination of Two~FiqInstitutions, California Institute of Tecnology 115

16. Palfrey, Thomas R., Multiple-Object, Discriminatory Auctions WithBidding Constraints: A Game-Theoretic Analysis, Social ST~§n§&-Working Paper 29, California Institute of ec nology, Oct(-Oer 1,14/1

APPENDIX A

SATELLITE AIRPORT PRO(IA1'

PiRPSED FACILITIES AND PROJEMI>.

MI-P

0 i)

o 2: z-4

-4UU0

LI W .4.6 0 .6.. 'a 'a I-~ 0~ .6

.6

.6 044 44 ,,..6h. ~ 4% 4%

U. ~ 8.6

-4 -4

bJ .64 .6h. .1%

a.6 .6

0 04

b C

U

* in(.0 ~ A

o *~ 6Ob .46 06

JE - ~.- S

- bg Ub

U~ 66.~.Lb b~Z NW -~ -4..

. U UM b~6s 0 60 -u* 44 *5W~4

ZUbIbUU 3.14 Ml MS6 *..~. .2 I~J44b 0.. 14436 ~ *U 64 6U 641 *2 24 -44

* .-. ~mu u.. N. 6. ~.. i. o. . . . -4 4 - i b

.4 S~~.. ..E - 3 3 0a .1

a

I6

:1 ii II

I.

)aM

-. 0U.'I o C * .~

4 I- * -4 p~ -. ~

=

CC. '0 4~f* 4 p~ ,-. '0 f~44'0 4 E~4 - - ~ -'

I __ ________________________________________________________

4

U * C

bE a4 6 0 U)

- .. ia. -. 4- ax

* - M -

I -u£ I 3 3~

~i I ~i ~ I :1~ U

j* I I* a* j~.1.ii I~

a.*1 a I U

-~f I I *1~aI

- S

1> .5

* C4* U

.5 .4.5 'a

.0b 1 1'

a,

I* a4' 0

L ~ *44' ~q* a a 2* b~ - -U U50 *N ;f. ~g

b .5 p.U Z..2 MU~ ~ 4~ 4* ~'2~j~ ~

I ~ I

U4 1.1

U

S. - -. .-

1

2.0 a I I

- 'a6

0I..

* 0 'a .~ a.' 9..4 9. 4 a.- 49 .9a. s.~ a.. - a.' a.- aS

ha $*..S '0 49 a.-,..4 '9.. '-4 4 '0 04 9.44.9a. 49 '0 - 49

9.4949

a.'

'6a.

49C I I

ml.4a.

A ~..-a. - I-~ a* ,..a. 16 -

.4 66.4 .4

U ii 8 A :a.2 -~ Ua :35 - 38~ a

- ~6 *

- I~ i~I.I.~ 2~ ~AI --

J [i*i** Z a. 0.3

U6'.

aI.

*1* -

5 83- ml * aU 6

- - - Ua.

~1 £ i I ml

z

I

8I.-

0b.

U ~ -E 'a4 .0 .0 4, 4 4% - 0a .0 .e .- 4,

a

a

~t.1 4%.49 .~ 04

49 49

I I* .4

2

~am3 4,4b. EE .4'

49

0

d ~,

C0

* - .4

0.4 U M

* * 4..a .4

U - £0I - - a.4 1

-U6. ~3 -

- - U i'~ .?* * ~* *a U ~-. I. -" .4

U. A ~ 'U. - -.4 0 -~ .4 ~ = U

.. a -~OMt.~ * 2 -* *q ~aq 0.u-~.4 .4 .4 ~. -

U.] a~ Ii j2~ !i~I..,. .1. 0 6 4' U

4, U U 1 .~ K -,

i~

00 6

~. i -

I .~ U-4Ii - -I

U U

6

Srd

4 a - C C ~'.6 - f~ t4 4 F- F- * Sb. 4 ~ 4 F- 4 4.4 4

4.4

a

a

U 4 @ 4 ~ C F-.6 - 4.4 4 4.~F- F-b. 4 ~ - 4 4.4 F-

5 4.4 4.4

*.6 F- 4 4b. 4.4 F- 4 4

S

.6U

-S i6

A =

I: .i 1.. - .4

U

-~ I.~i I. ..aq WE *~6.4 6~ .J~ ~ q.4 * EU UI ~I. 0.4W a UUU ~* A W 6.. 0. 0.. a~ * 5.. ... h.- U * 6 6 .4 U * * bI: ~ I ~ I

U I Ius.

U I

- U

S I 11112

I

F.

40 .0-

b. '0

8pim 106

~~ba

ob M - -

w & WN 0% W%-

a.1 r 4'

46

a.0

- so

" 5

4 -. q

6Ut +,

'U-

6. -]i +i i~ a

'U ~ a:! 'i4-6+: j. ! °

_; , ~ ~ -.. j j .

i, J-•

-

-0

~0

InIU 0

4%i

* 9-