high density en route airspace safety level and … · tma precision approaches crm ils (1980)...

HIGH DENSITY EN ROUTE AIRSPACE SAFETY LEVEL AND COLLISION RISK

ESTIMATION BASED ON STORED AIRCRAFT TRACKS

(EIWAC 2010)

+E. Garcia*, F. Saez**, R. Arnaldo*** CRIDA (ATM Research, Development and Innovation Reference Centre)

Madrid, Spain

[email protected]

**Air Navigation Department

Universidad Politécnica de Madrid (UPM)

Madrid, Spain

[email protected]

[email protected]

CRIDA. CRM 3D. ENRI Int. Workshop on ATM/CNS. Tokyo, Japan. (EIWAC 2010) 11/10/2010 2

CRIDA an Initiative from Aena(ATM Research, Development and Innovation Reference Centre)

Scientific Rigour

Result Oriented

Flexibility

UniversitySpanish Air Navigation

Service Provider

Transport EngineeringCompany

CRIDA. CRM 3D. ENRI Int. Workshop on ATM/CNS. Tokyo, Japan. (EIWAC 2010) 11/10/2010 3CRIDA. CRM 3D. ENRI Int. Workshop on ATM/CNS. Tokyo, Japan. (EIWAC 2010) 11/10/2010 3

Outline

• INTRODUCTION

• NEED FOR A NEW COLLISION RISK MODEL

• BACKGROUND

• 3D CRM – GOAL

• 3D CRM – GENERAL DESCRIPTION

• RADAR DATA PROCESSING

• SAFETY METRICS

• CONCLUSIONS AND WORK IN PROGRESS


INTRODUCTION

EUROCONTROL has worked in the last years to develop the 3-D collision risk model (CRM).

The 3-D collision risk model was developed as a general mathematical framework to assess the level of safety in continental en-route airspace, where controllers monitor air traffic by means of radar surveillance and provide aircraft with tactical instructions

ATCO

The objective of the software prototype tool is not only to eventually produce an estimation of the level of safety achieved in the airspace under assessment but also to provide safety-related metrics and trends, which can be monitored over time.


NEED FOR A NEW COLLISION RISK MODEL

Previous models no not appropriate assess and monitor the level of safety in highdensity en-route radar airspaces using as a sole source of input data therecorded aircraft trajectories. Traditional approaches to Collision Risk Models (CRM), do notcapture the complexity inherent to an operational radar environment like the one in Europe,with high amount of traffic, a large number of crossings tracks, climbing and descendingaircrafts and complicated route structure.

PHASE MODELO

EN-ROUTE

Airway structureconsisting of one ormore parallel routes.No radar surveillance.

CRM Reich (1960)

Improvement ofCollision Risk Models

Extended CRMReich (1993)

Parallel routes.Surveillance base donADS.

CRM ADS (1993)

TMA

Precision approaches CRM ILS (1980)

Landing on closely andultra closely spacedrunways

BRM (1990)

MIT (1997)

Future operationalconcepts

NLR (2001)


BACKGROUND

Nowadays, ANSP and Civil Aviation Authorities (CAA) mainly use ATM accident and incident databases to monitor and provide evidence of levels of safety. However, although these databases are very powerful tools and are improving constantly, they still have some weak points that need to be considered:

•Not all incidents are reported by pilots and air traffic controllers. In fact, it is very difficult to infer how many real incidents have occurred for each one that is reported.•Incident severity is generally ranked solely on how close aircraft get, without considering the geometry of the event or other parameters, e.g. closure rate.•Incident Classification is not homogeneous in all databases. Furthermore, special care has to be taken to train database personnel so that the same classification criteria always apply.


BACKGROUND

MANAGEMENT AND ANALYSIS

OF INFORMATION

Aircrafttrajectories fromSimulationsE.gg: RTS, FTS,…

Aircraft PerformancesE.g.. BADA,…

ATC Procedures

Airports DataE.g.: AIS,…

Airspace informationE.g. AIS, Capacity, Demand

Meteo InformationE.g.: METAR,……

Aviation accidents and Incidents

Aviation MaintenanceIncident Database

Real Aircraft TrajectoriesE.g. :radar data, ADS,…

CRM 3DANALYSIS,

ASSESSMENT AND MONITORING OF

OPERATIONAL SAFETY


3D CRM – GOAL

A major objective of the 3-D CRM tool is to complement the information collected in the accident and incident databases, thereby providing:

1. Identification of all proximate events based on radar data.

2. Complete classification of all proximate events using clear and consistent criteria.

3. Detailed information on the evolution of each proximate event.

4. Collision risk estimate


3D CRM – GENERAL DESCRIPTION

AirspaceScenario

3D CRM Analytical

Model

Safety Level

Estimation

3D CRM Radar Analysis Tool

Traf

fic

Sam

ple

Safety Event Reports


3D CRM – GENERAL DESCRIPTION

AirspaceScenario

Radar Data Files

Identify Proximate

Events

Analyse and Classify Events

Probability of Risk

Other Safety Metrics

Segmentation

Coarse Filter

Radar Data Processing Module Parameter Estimation Module


RADAR DATA PROCESSING

The radar data processing module comprises two different : •decoding and storage of radar track and flight plan files, and •track segmentation of radar data.

-3 -2 -1 0 1 2

x 105

-1

0

1

2

3

4

5x 10

5

X

Y

-3 -2 -1 0 1 2

x 105

-1

0

1

2

3

4

5x 10

5

X

Y

X (m) X (m)

Y (

m)

Y (

m)



GOAL ->Identify Proximate Events:

R

2 HConflict:Aircraft physical separation< intended margin (D, H)

Collision: Aircraft physical separation< aircraft physical sizes

Conflicts

Potential Conflicts

Potential Collisions



Look-ahead time (LAT): is the time horizon within which all aircraft positions are projected to explore existence of “potential conflicts/collision”.

-> 10 minutes

CPA

D*



MAP of HOT SPOTS


SAFETY METRICS ESTIMATION

Analysis of Proximate Events

Classification of Proximate Events

Detection of Activated Alert Systems

Identification of ATM System Weaknesses


tCI

v1

v2

CPA(t cpa ,[r12]cpa,[h12]cpa)

t cpa : time at CPA

[r12]cpa: horizontal separation at CPA

[h12]cpa: vertical separation at CPA

[r12]cpa<R & [h12]cpa<H

tCI : time of identification of potential

conflict

SAFETY METRICS ESTIMATIONAnalysis of Proximate Events


tCI

v1 *

v2

t cpa *: new time at CPA

[r12]cpa *: new horizontal separation at CPA

[h12]cpa *: new vertical separation at CPA

[r12]cpa * <R & [h12]cpa * <H

t m: time of maneuvre

Change in the Rate of Climb (ROC)

New CPA*(t* cpa ,[r12]*cpa,[h12]*cpa)

tm



tCI

v1

v2

[h12]cpa : vertical separation at CPA

[h12]cpa >H

t m: time of a conflict resolution manoeuvretm tCR

Climb->Level



tCI

H

r12: actual horizontal separation

h12: actual vertical separation

r12 <R & h12 =H

t m: time to entry into conflict volumetm tCF




NATUREVERTICAL REGIME

RELATIVE HEADING

A/C REACTIONTRAFFIC TYPE

Passing Event

PROXIMATE EVENT

Conflict

Potential Conflict

Level-Level (LL)

Climb-Level(CL)

Descend-Level(DL)

Climb-Climb(CC)

Climb-Descend(CD)

Descend-Descend(DD)

Same

Opposite

Crossing

Civil Traffic

Both Military Aircraft

One Military Aircraft

CRITERIA

Change Vertical Profile

Modify Heading

ACTIVATED ALERT SYSTEM

TCAS TA

STCA

TCAS RA

Change of Speed

No Alert No Reaction Potential Collision

False Detection

Classification of Proximate Events



[r12]CPA

tCI

Aircraft 2

Aircraft 1

tCF tCPA

tCI: identification of potential conflicttCR: start of a conflict resolution manoeuvretCF: predicted time of entry into conflict volumetCPA: predicted time of CPA[r12]cpa: horizontal separation at CPA

t2 :Time to Conflict

t3 :Time-to-go to CPA

R

Predicted trajectory

t1 : Overall Reaction Time

t1=tCR-tCI

t2=tCF-tCI

t3=tCPA-tCI

tCR




35 sec48 sec

2 mins4 mins

TIME TO CPA

Potential Collision

TCAS RA(See & Avoid,..)

TCAS TATACTICAL

Safety Barriers

35 sec48 sec2 mins4 mins

TIME TO CPA

( )P Tactical Failure

( )RiskP

( )P STCAFailure

( )P TCASTAFailure

( )P TCAS RAFailure

No lossof Sep.

No loss of Sep.

Loss ofSeparation?

Loss ofSeparation

Collision

Yes

No

Yes

No

Yes

No

Yes

No

STRATEGIC

Yes

No(Pre )P Tactical Failure

TACTICALBASED ON STCA


No lossof Sep.



VERTICAL CLOSURE RATE (fpm)

Strategic

ATC Action TacticalATC Action(STCA)

Traffic Advisory (TA)

35 sec

VE

RT

ICA

L S

EPA

RA

TIO

N (f

eet)

48 sec

2 min

H

850 600

ResolutionAdvisory (RA)



SAFETY METRICS ESTIMATIONDetection of Activated Alert Systems

HORIZONTAL PLANE VERTICAL PLANE

-5 -4.5 -4 -3.5 -3 -2.5 -2

x 105

-1

-0.5

0

0.5

1

x 105

X

Y

204.0 kts

22.6 22.65 22.7 22.75 22.8 22.85 22.9 22.95 23 23.05

320

330

340

350

360

370

380

time (h)

Alt

itu

de

(F

L)

0.0 fpm

1875.0 fpm

RARA

Alt

itud

e(F

L)

Y

Time (h)


Proximate Event Data: AC 1 - AC 2

[r12]cpa

(NM)

[h12]cpa

(ft)

TTC

(min)

Dur

(min)

TCPArel

ManoeuvreAC1

ManoeuvreAC2

Legend

T0 0.66 -0.00 5.99 0.001.67

-0.42 00.00 -5.12 -0.00

T1 0.24 0.00 0.86 -0.005.12

>Conv(85º)<

Turn 5.7º

TCAS RA

0.01 -0.00 -0.40 -0.00

T2 0.25 0.00 0.46 -0.000.40

AC1 Turn ; ><

CL FL

CONFLICT RESOLVED-VERTICAL PLANE

0.11 1000.00 -0.23 0.00

T3 0.36 1000.00 0.22 0.000.24

AC2 CL FL ; VER SEP; >Conv(88º)<

MIN REAL SEP: 0.3 NM 1000 ft

-0.01 00.00 -0.22 -0.03

T4 0.34 1000.00 0.00 -0.030.48

>Conv(93º)<

2.44 -0.00 0.00 0.29

T5 2.79 1000.00 0.00 0.260.80

<Div(99º)>

Tm


0 100 200 300 400 500 600 700 800 9000

5

10

15

20

25

30

35

40

45

50

Horizontal Closure rate (kts)

Ho

rizo

nta

l D

ista

nce

(n

m)

Resolution Advisory

Traffic Advisory

Tactical ATC Action (STCA)

Strategic ATC Action

T0

T1

T3

T4

0 1000 2000 3000 4000 5000 60000

1000

2000

3000

4000

5000

6000

7000

8000

Vertical Closure Rate (fpm)

Ve

rtic

al S

ep

ara

tio

n (

ft)

Resolution Advisory

Traffic Advisory

Tactical ATC Action (STCA)

Strategic ATC Action

T1

T3T4

As an example, this figure represents in the TCAS and STCA diagram the evolution of the encounter described beforte. A TCAS RA is activated if the kinetic and geometric characteristics of the event in the horizontal plane and in the vertical plane are in the red area at the same time.



HAZARD(potential collision)

Traffic Density

+

Traffic complexity

+

ATC/Pilot Errors

+

Technical Failures

Initiating events leading

to potential collisions

Detection / Resolution of

potential collisions

ATCO STCA TCAS …

No Safety Effect

Conflict

Collision

Risk Context Qualification of HazardsEffectiveness of Safety

Barriers




Risk Context (within

selected area/time-

frame)

Effectiveness of

Safety Barriers

Flight-Time Hot Spots Number of Real Conflicts

Number of MovementsClassification of Proximate

Events:

• Nature

• Traffic Type

• Vertical regime

• Relative heading

• A/C Reaction

• Activated Alert System

(TCAS, STCA)

Overall Reaction Time

Kinematics

(Speed, Type of Aircraft)Time Margin

Nº of Entries and Exits Time to Conflict

Route structure Time-to-go to the CPA

Traffic Density

% of Potential Conflicts

solved in:

• vertical plane and

• horizontal plane

Number of Routes

Correlation between

Hot Spots and

Traffic Density Maps.

Average Nº of Aircraft “near”

a Proximate Event

% Evolving (non-level) Aircraft Nº of potential collisions

Separation Rules

Qualification of Hazards




Side Length=7.5 NM

Risk Context

(within selected

area/time-frame)




Qualification of

Hazards

-4 -3 -2 -1 0 1 2

x 105

-2

-1

0

1

2

3

4

x 105

CIVIL-MILITARY TRAFFIC

MILITARY TRAFFIC

CIVIL TRAFFIC

Total Number of Proximate Events: 45389

CIVIL TRAFFIC 91.2%

MILITARY TRAFFIC 3.2%

CIVIL-MILITARY TRAFFIC 5.6%


POTENTIAL CONFLICT 84.86%

ACTUAL CONFLICT 0.08%

FALSE DETECTION 14.59%

PASSING EVENT 0.46%

LEVEL-LEVEL 13.8%

LEVEL-CLIMB 59.6%

LEVEL-DESCENT 7.1%

CLIMB-CLIMB 11.9%

CLIMB-DESCENT 6.4%

DESCENT-DESCENT 1.2%


-4 -3 -2 -1 0 1 2

x 105

-2

-1

0

1

2

3

4

x 105

TCAS RA



Effectiveness of Safety Barriers

ATC STRATEGY 69.6%

STCA 28.0%

TCAS TA 2.1%

TCAS RA 0.3%

0 1 2 3 4 5 6 7 8 90

1

2

3

4

5

6

7

8

9

10

Overall Reaction Time (min)

Tim

e-t

o-g

o t

o t

he C

PA

(m

in)

0

10

20

30

40

50

60

70

80

90

0 1 2 3 4 5 6 7 8 9 100

100

200

300

400

500

600

700

800

900

Overall Reaction Time (t1)

Overall Reaction Time (min)

0 1 2 3 4 5 6 7 8 9 100

100

200

300

400

500

600

700

Time-to-go to CPA (t3)

Time-to-go to CPA (min)


CONCLUSIONS

The 3-D CRM tool has been designed to complement the information collected in the accident and incident databases, thereby providing the following information inferred from the in depth assessment of proximate events:

Identification of all proximate events based on radar data.

Complete classification of all proximate events using clear and consistent criteria.

Detailed information on the evolution of each proximate event.

Safety metrics and other air traffic factors


WORK IN PROGRESS

•Apply the 3-D CRM too to traffic samples of different airspaces and extend the principles of 3-D CRM from en-route to Terminal Manoeuvring Area (TMA) scenarios;

•Develop a methodology to provide a complete risk picture of the scenario, identifying the ATM system weakness and characterizing the performance of the safety barriers, using all the information provided by the 3-D CRM tool that could be used by ATM service providers to monitor and improve safety levels in their operation;

•Complete an analytical model based on the 3-D CRM tool to provide true collision risk values.

CRIDA. CRM 3D. ENRI Int. Workshop on ATM/CNS. Tokyo, Japan. (EIWAC 2010) 11/10/2010 34CRIDA: General Description

CRIDA (Centro de Referencia I+D+i ATM)

Questions

high density en route airspace safety level and … · tma precision approaches crm ils (1980)...

Documents