for individual and collective training · for individual and collective training larry l. meliza...

AD-A242 753Technical Report 936 \l~ll\IlIl\lI\\ll\ a

Early Training Strategy Developmentfor Individual and Collective Training

Larry L. Meliza and Bruce W. KnerrU.S. Army Research Institute

August 1991

United States Army Research Institute_- Ifor the Behavioral and Social Sciences -

wi 0 0

Approved for public release; distribution is unlimited.

U.S. ARMY RESEARCH INSTITUTE

FOR THE BEHAVIORAL AND SOCIAL SCIENCES

A Field Operating Agency Under the Jurisdictionof the Deputy Chief of Staff for Personnel

EDGAR M. JOHNSON MICHAEL D. SHALERTechnical Director COL, AR

Commanding

Technical review by

Billy L. BurnsideMichael J. Singer - ....

W t ., A 1 ,t

. . . . . . . . . . . . .. . '. ,. u

! 1'~

NOTICES

DISTRIBUTION: Primary distribution of this report has been made by ARI. Please addresscorrespondence concerning distribution of reports to: U.S. Army Research Institute for theBehavioral and Social Sciences, AT'N: PERI-POX, 5001 Eisenhower Ave., Alexandria, Virginia22333-5600.

FINAL DISPOSITION: This report may be destroyed when it is no longer needed. Please do notreturn it to the U.S. Army Research Institute for the Behavioral and Social Sciences.

NOTE: The findings in this report are not to be construed as an official Department of the Armyposition, unless so designated by other authorized documents.

UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE

RE R DForm ApprovedREPORT DOCUMENTATION PAGE OMB No. 0704-0188

la. REPORT SECURITY CLASSIFICATION lb. RESTRICTIVE MARKINGS

Unclassified --2a. SECURITY CLASSIFICATION AUTHORITY 3. DISTRIBUTION IAVAILABILITY OF REPORT

-- __Approved for public release;2b. DECLASSIFICATION /DOWNGRADING SCHEDULE distribution is unlimited.

4. PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S)

ARI Technical Report 936

6a. NAME OF PERFORMING ORGANIZATION 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATIONU.S. Army Research Institute (If applicable) --

PM TRADE Field Unit PERI-IF

6c. ADDRESS (City, State, and ZIP Code) 7b. ADDRESS (City, State, and ZIP Code)

12350 Research Parkway

Orlando, FL 32826-3276

Ba. NAME OF FUNDING/ SPONSORING 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION U. S. Army Research (If applicable)

Institute for the Behavioral --and Social Sciences PERT-I8c. ADDRESS (City, State, and ZIP Code) 10. SOURCE OF FUNDING NUMBERS

PROGRAM PROJECT [TASK JWORK UNIT5001 Eisenhower Avenue ELEMENT NO. NO. NO. ACCESSION NO.Alexandria, VA 22333-5600 62785A 790 3105 Hl

11. TITLE (Include Security Classification)

Early Training Strategy Development for Individual and Collective Training

12. PERSONAL AUTHOR(S)

Meliza, Larry L.; and Knerr, Bruce W.

13a. TYPE OF REPORT 13b. TIME COVERED 14. DATE OF REPORT (Year,Month,Day) 1l. PAGE COUNTFinal J FROM 89/10 TO 90/08 1991, August

16. SUPPLEMENTARY NOTATION

17. COSATI CODES 1 8jB. SUBJECT TERMS (Continue on reverse if necessary and identify by block number)FIELD GROUP SUB-GROUP ,ptimization of simulation-based training systems)(OSBATS)I , Automated systems approach to training)(A&AT

Embedded training ( o (Continued)

19, ABSTRACT (Continue on reverse if necessary and identify by block number)The training strategy for a new weapon system identifies the training devices required,

the tasks each device will be used to train, and the circumstances under which each device

will be employed. Consideration of embedded training (i.e., use of operational equipment and

training software to provide training) as the first option for new weapon systems forces

early development of training strategies. Training development tools, such as the Optimiza-

tion of Simulation-Based Training System, are available to support development of a training

strategy, but an overall model is needed to show how the various tools can be integrated to

support strategy development. This report describes a high-level model for early training

estimation that incorporates other training development tools. The benefits of this model

include integration of individual skills training across duty positions, individual skills

training with collective training, collective task training across unit missions, and col-

lective task training across echelons. j V

20. DISTRIBUTION /AVAILABILITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATIONrUNCLASSIFIED(UNLIMITED 0l SAME AS RPT. 0l DTIC USERS Unclassified

22a. NAME OF RESPONSIBLE INDIVIDUAL 22b, TELEPHONE (Include Area Code) 22c. OFFICE SYMBOL

Larry L. Meliza 1 (407) 380-4374 PERI-IF

DD Form 1473, JUN 86 Previous editions are obsolete. SECURITY CLASSIFICATION OF THIS PAGEUNCLASSIFIED

UNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE(When Data Entered)

ARI Technical Report 936

18. SUBJECT TERM4S (Continued)

T-rminmirtg- -davices"Training strategy.

UNCLASSIFITEDSECURITY CLASSIFICATION OF THIS PAGE(*%*n Date Entered)

ii

Technical Report 936

Early Training Strategy Developmentfor Individual and Collective Training

Larry L. Meliza and Bruce W. KnerrU.S. Army Research Institute

ARI PM TRADE Field Unit at Orlando, FloridaStephen L. Goldberg, Chief

Training Research LaboratoryJack H. Hiller, Director

U.S. Army Research Institute for the Behavioral and Social Sciences5001 Eisenhower Avenue, Alexandria, Virginia 22333-5600

Office, Deputy Chief of Staff for PersonnelDepartment of the Army

August 1991

Army Project Number Human Performance Effectiveness20162785A790 and Simulation

Approved for public release, distribution is unlimited.

iii

FOREWORD

The Army's need for a top-down approach for developingtraining to support new weapons systems was reported by the ArmyScience Board in 1985. In response to this need, the U.S. ArmyResearch Institute for the Behavioral and social Sciences (ARI)is conducting research to develop and evaluate training designrules and guidelines that are applicable early in the weaponssystem design process. These guidelines will facilitate devel-opment of an integrated set of requirements for training devices,simulators, and simulations, including embedded training, forboth weapons systems and units.

This report documents the results of the initial work inthat process. The first goal was to create a preliminary modelof the process by which the training strategies for new weaponssystems are developed. The second goal was to estimate thebenefits of implementing the model. The final goal was toidentify the research required before the model could beimplemented.

The work described in this report is part of research task3105, Techniques for Early Estimation of Training System Require-ments, conducted for the Army Project Manager, Training Devices(PM TRADE), by the ARI PM TRADE field unit under a Memorandum ofUnderstanding, "Expanded MOU Between PM TRADE and ARI," dated 14July 1986. The model described in this report provides the basisfor development of a more detailed systems engineering analysisof the training development process.

EDGR JHNSONTechnical Director

v

EARLY TRAINING STRATEGY DEVELOPMENT FOR INDIVIDUAL AND COLLECTIVE

TRAINING

EXECUTIVE SUMMARY

Requirement:

The training strategy for a new weapon system identifies thetraining devices required, the tasks each device will be used totrain, and the circumstances under which each device will beemployed. The Department of the Army requirement to considerembedded training (i.e., use of operational equipment and train-ing software to provide training) as the first option for newweapon systems forces early development of training strategies.

Training development tools are being developed that cansupport parts of the process of developing a training strategy.The Automated Systems Approach to Training (ASAT) is designed tosupport preparation of the Army Training and Evaluation Program(ARTEP) documents that guide the training of units on collectiveand individual tasks (Science Applications International Corpora-tion, 1988). The Optimization of Simulation-Based Training Sys-tems (OSBATS) is a collection of prototype tools that assist inthe design of training devices (Sticha, 1988; Sticha, Blacksten,Buede, Singer, Gilligan, Mumaw, and Morrison, 1988). What ismissing is an overall model that shows how these and other toolscould be integrated to support training strategy development. Aneed exists to create a preliminary model of the process by whichthe training strategies for new weapons systems are developed,estimate the benefits of implementing the model, and identify theresearch required before the model could be implemented.

Procedure:

This report describes an early training estimation model interms of four functions: ARTEP development, front-end analysis,training strategy development, and training strategy comparison.The first two functions address identifying tasks to be trainedand analyzing training-relevant features of these tasks. Thethird function addresses generating feasible training strategyoptions. The fourth function is concerned with comparing thecost effectiveness of training strategy options. Each functionwas defined in terms of goals, subfunctions, inputs, outputs,benefits, and developmental risks.

vii

Findings:

Major benefits that accrue from this model are listed below.

* It focuses the training device design process on pre-paring units to perform their missions, the bottom lineof Army training.

* It allows for comparing the cost effectiveness of embed-ded training with other options very early in weaponsystem development.

• It integrates individual skills training across dutypositions, individual skills training with collectivetraining; collective task training across unit missions;and collective task training across echelons.

* It identifies individual tasks with a high payoff interms of progressive training value.

" It defines collective training targets of opportunitywith a high payoff in terms of progressive trainingvalue.

* It ensures minimal duplication among training devices andproducts in terms of training objectives addressed.

* It provides the level of detail regarding fidelity andinstructional support feature requirements needed tocompare the cost effectiveness of device alternatives.

" It ensures that embedded training is used when it is themost cost effective option for individual or collectivetraining.

Implementation of the concept required to gain the benefitslisted above will require the development and integration of newtools to address collective training, unit training, and newertraining device options like networked simulation.

Utilization of Findings:

These findings will be used by the Project Manager forTraining Devices (PM TRADE), the U.S. Army Training DoctrineCommand (TRADOC), and the U.S. Army Research Institute for theBehavioral and Social Sciences (ARI) to define future developmentof training design tools in a way that ensures they will supportthe development of integrated training strategies.

viii

EARLY TRAINING STRATEGY DEVELOPMENT FOR INDIVIDUAL AND COLLECTIVETRAINING

CONTENTS

Page

INTRODUCTION ............. ....................... 1

Background ............. ....................... 2Objectives ............. ....................... 6

EARLY TRAINING STRATEGY MODEL ......... ............... 7

Overview and Benefits of the Model ...... ........... 7ARTEP Development Function ........ ............... 10Front-End Analysis Function ..... .............. 14Training Strategy Generation Function .. ......... 18Training Strategy Comparison Function .. ......... 26

WORK REQUIRED TO IMPLEMENT MODEL .... ............. 31

SUMMARY ............ .......................... 33

REFERENCES .......... ........................ 35

LIST OF FIGURES

Figure 1. Overview of the ideal process for developingtraining strategies for a new weapon system 3

2. Early Training Strategy Development Modeloverview ............ ................... 8

3. ARTEP function for estimating the effectsof a new weapon on unit training requirements 11

4. Front-end analysis function for developingtraining-relevant information about tasks . . . 15

5. Training strategy generation function fordeveloping feasible and practical trainingstrategy options ...... ............... 19

6. Training strategy comparison function forestimating relative cost effectiveness oftraining strategies ..... ............. 27

ix

EARLY TRAINING STRATEGY DEVELOPMENTFOR INDIVIDUAL AND COLLECTIVE TRAINING

Introduction

The goal of training Army units to perform their missions isbecoming more difficult to attain as both the cost of training onoperational equipment and the need to reduce expendituresincrease (Armstrong and Deaver, 1990). The careful applicationof training devices and simulators to new weapon systems is oneway to address this problem. Devices and simulators can make itpossible to provide soldiers with hands-on practice without theexpenditures for ammunition, fuel and equipment maintenancetypically associated with training on operational equipment.

Today a wide range of media are available that can be usedto supplement classroom lectures and training on operationalequipment to meet training needs. These media range in complexityfrom the networking of full-mission simulators that simulate allor most of the subsystems of an item of equipment to forms ofcomputer-based instruction which are knowledge-based and involveno simulation. The plan or strategy for applying training devicesto support training on new weapon systems is developed throughthe process of concept formulation to ensure that training isconducted in a cost-effective manner (U.S. Army Materiel Commandand U.S. Army Training and Doctrine Command, 1988).

The training device strategy for a new weapon systemidentifies the training devices required, the tasks each will beused to train, and the circumstances under which each device willbe employed (U.S. Army Training Support Center, 1988). Thisstrategy must of course be part of a total training strategythat includes training methods in addition to those that aredevice-based. The training strategy should describe "theintegration of the training subsystem and the development of thetotal [weapon] system and the integration of the developingsystem into ongoing training systems" (U.S. Army Training andDoctrine Command, 1982).

In addition to meeting the requirements noted above, thetraining strategy for a new weapon system must be developed earlyin the development of the weapon system. According to TRADOCRegulation 351-9 (1982), the training strategy or OutlineIndividual and Collective Training Plan (OICTP) should "identifythe constraints which training requirements and resources mayimpose on the design of the materiel system". The option ofembedded training (ET), "training that is provided bycapabilities designed into or added to operational systems"(Department of the Army, 1987), increases the importance of anearly integration of training strategies and weapon systemdevelopment (Strasel, Dyer, Roth, Alderman, and Finley, 1988.)

1

The 1985 Army Science Board Summer Study on Training andTraining Technology reachpd certain conclusions relevant to theproblem of strategies fo- training devices (Peden, Barth, Bonder,Caro, Fried, Jones, LaBerge, Morrison, O'Neal, Pauly, Pettigrew,Rathjen, Simmons, Welch, Williges, and Zarafonetis, 1985).First, the study concluded that "a top-down systems engineeringapproach to the definition and development of training systemsfor units is not being accomplished" and "a total systemstechnique will yield the most efficient mix of devices, ranges,simulators and simulations in an effective manner." Second, thestudy concluded that "current training guidance focuses mostly onindividual devices, simulators, and simulations" and there is aneed for top-down training strategies which incorporate andintegrate all appropriate training devices.

The Army would benefit from a systematic means of developingand assessing training strategies early in concept formulationfor new weapon systems. This would ensure timely and accurateconsideration of embedded training and enhance the overall cost-effectiveness of the training system. This systematic methodshould include individual training, collective training,institutional training and unit training.

Background

The functions performed by the training design community indeveloping and refining strategies are summarized in Figure1. The functions of identifying the tasks to be trained,analyzing tasks and developing training strategy options are theresponsibility of Army school training developers (U.S. ArmyTraining Support Center, 1989). The training strategy optionswhich emerge from this portion of the process should reflectlessons learned when analyzing the training-relevantcharacteristics of the task including the conduct of a mediaanalysis. The job of developing information to be used incomparing strategy options is a cooperative effort of ArmySchools and the U.S. Army Project Manager for Training Devices(PM TRADE) referred to as the Trade-Off-Determination(Department of the Army,1989). The TRADE-Off-Determinationincludes such considerations as the technical feasibility ofvarious options and the cost of implementing these options.Comparison of training strategy options to select thebest option is a joint offort of Army schools and PM TRADEand the analyses which combine to prove the best mix of trainingdevices are referred to as the Trade-Off Analysis and the Costand Training Effectiveness Analysis (U.S. Army Materiel Commandand U.S. Army Training and Doctrine Command, 1988).

2

IDENTIFY TASKSTO BE TRAINED

ANALYZE TASKS

DEVELOP TRAININGSTRATEGY OPTIONS

DEVELOP INFORMATIONFOR USE IN COMPARINGSTRATEGY OPTIONS

COMPARE STRATEGYOPTIONS TO SELECTMOST COST-EFFECTIVEOPTION

Figure 1. Overview of the ideal process for developing trainingstrategies for a new weapon system.

Problems in developing and comparing training devicestrategies have been described by various researchers. First,training devices are often selected before the tasks to betrained are identified and analyzed (Heeringa, Baum, Holman, andPeio, 1982; Dynamics Research Corporation, 1984, Meliza andLampton, in preparation). This problem is due, in part, to thefact that training device design and acquisition is often drivenby technology rather than training strategies, and the trainingrequirements addressed by the device are often determined by thecapabilities of a specific training technology rather than therequirements of the task or tasks to be trained (Hofer, Ozkaptanand Kincaid, 1987). That is, Army schools often adopt aparticular training device option without first finding outwhether the option is truly cost-effective. Second, trainingstrategies are often vague and fail to link the tasks to betrained with specific training devices (Meliza and Lampton, inpreparation).

3

The Army recognized the need to develop procedures forproducing quick, early estimates of the training resourcerequirements necessary to provide training on new weapon systemsat least a decade ago (Roth, Warm,Peters, O'Brien, Hawley, Pence,Robinson, Masterson, and Criswell, 1989). In 1984, DynamicsResearch Corporation noted that "Early front-end analysistechniques which could help produce more objective estimates oftraining requirements early in the acquisition process, are notemployed in OICTP (Outline Individual and Collective TrainingPlan) preparation". A subsequent Army Science Board Ad HocSubgroup on Army Analysis concluded that front end analyses arenot being performed consistently (Christie, 1987).

MANPRINT Product Four (Roth, Warm, Peters, Masterson andCriswell, 1987; Ditzian, Roth and Johnston, 1987) is the mostrecent in the HARDMAN series of tools for producing manpower,personnel and training resource estimates (Mannle, Guptill, andRisser, 1985). It is concerned with improving the quality ofearly estimates of training resource requirements for new weaponsystems by identifying weapon systems and subsystems already inexistence which might be expected to have training requirementssimilar to those of a new weapon system. MANPRINT Product Fourgives training developers access to data on the current trainingstrategy for these comparable systems. MANPRINT Product Four byitself is not a complete solution to the problem of earlytraining estimation because it specifically addresses onlyindividual task training within Army schools. Collectivetraining and individual training in units are not addressed byMANPRINT Product Four.

The primary focus of MANPRINT Product Four is to developestimates of the resources required to conduct training on a newweapon system based on resource requirements for conductingtraining for comparable systems. This approach is of limitedutility as the Army attempts to develop training strategiesencompassing individual and collective training in institutionsand units. The proliferation of training device options (such asET and simulation networking) combined with the increasedemphasis on applying devices to individual, collective andcombined arms training in units has created a large gap betweenearly training estimation techniques that are available versusthose that are needed.

Training Development Tools. In an attempt to incorporatecollective training, unit training and new device options (suchas ET) into early training estimation, ARI and the U. S. ArmyProject Manager for Training Devices (PM TRADE) sponsored thedevelopment of an architectural model for an Integrated TrainingSystem Decision Support System, referred to as TRASER. TheTRASER model (Hinton, Feuge, Braby, Stultz, Evans, Gibson, andZaldo, in preparation) describes the training development processfor a new weapon system throughout the four phases of the Life

4

Cycle Systems Management Model (LCSMM). In addition toincorporating collective and unit training into the area of earlytraining estimation, TRASER attempts to integrate varioustraining development tools. These tools include MANPRINT ProductFour, and three other systems under development within ARI; theAutomated Systems Approach to Training (ASAT), the Optimizationof Simulation Based Training Systems (OSBATS) and EmbeddedTraining.

ASAT is designed to aid schools in preparing the ArmyTraining and Evaluation Program (ARTEP) documents whichguide the training of units on collective and individual tasks(Science Applications International Corporation, 1988). ASAT(Bloedorn, Crooks, Merrill, Saal, Meliza and Kahn, 1985) wasintended to reduce the work required to revise the ARTEPdocuments required to field new weapon systems and produce ARTEPMission Training Plan (AMTP) and Drill documents. Thesedocuments provide such information as how to integrate thetraining of specific individual and collective tasks.

OSBATS is a decision support system which uses informationabout specific individual tasks, device options, instructionalsupport features, and student throughput to compare the cost-effectiveness of training device design options for individualtraining in institutions (Sticha, Blacksten, Buede, Singer,Gilligan, Mumaw and Morrison, 1988). OSBATS employs five modules(Simulation Configuration, Instructional Feature Selection,Fidelity optimization, Training Device Selection, and ResourceAllocation) which can be used in an iterative fashion torefine design options.

Current ET guidance is in the form of a ten-volume set ofinstructions intended to aid training designers in identifyingcandidate systems or tasks for embedded training and subsequentimplementation of embedded training (Finley, Alderman, Peckham,and Strassel, 1988). Volume 2 of the ET guidance (Strasel, Dyer,Roth, Alderman and Finley, 1988) addresses the early selection ofET candidates by asking the four broad questions listed below.

- Are there general and/or weapon-specific policydecisions which dictate the use of embedded training?

- Do the tasks involved in maintaining and operating theequipment require sustainment training?

- Is it feasible to develop an embedded trainingcomponent for the weapon system?

- Is it likely that embedded training will be a cost-effective alternative to other training options?

5

Training development tools exist which can support parts ofthe process of developing a training strategy. What is missingis an overall model which shows how these and other tools ;ouldbe integrated to support training strategy development,especially early in the concept formulation process for trainingdevices. While the TRASER model makes many valuablecontributions to the design of an early training estimation andrefinement system, details about how the various trainingdevelopment tools would support the development of trainingstrategies are lacking. Further, the TRASER model integrates theabove training development tools within the phases of the LCSMMfor which they were originally designed. MANPRINT Product Fourand ET are largely in the first phase of the LCSMM, while ASATand OSBATS are in the second phase of the LCSMM. The quality ofearly training estimation should be enhanced by using ASAT andOSBATS as tools within the first phase of LCSMM.

The authors assume that the reader is familiar with thetools described above. The reader is referred to Hinton et al(in preparation) for information on TRASER, Bloedorn et al (1985)for information on ASAT, Ditzian et al (1987) for information onMANPRINT Product 4, and Sticha et al (1988) for information onOSBATS.

Objectives

One objective of this report is to present a preliminaryhigh level model for early development and comparison of trainingstrategies which incorporates ASAT and OSBATS by using data fromcomparable weapon systems. A second objective is to estimate thebenefits of implementing the model. The third objective is toidentify the research required before the model could beimplemented. The discussion of the model and the benefits beginson page 7 of the report. The description of the work requiredbegins on page 31.

The model presented in this report places little emphasis onthe development of strategies for classroom individual trainingin institutions, because other tools appear to address that areaadequately. The model includes training using operationalequipment or training devices within institutions, and itincludes individual and collective training in units.

6

Early Training Strategy Model

Overview and Benefits of the Model

The early training estimation model is described in terms offour functions; ARTEP Development, Front End Analysis, TrainingStrategy Generation, and Training Strategy Comparison (see Figure2). The first two functions address the activities ofidentifying tasks to be trained and analyzing training-relevantfeatures of those tasks. These functions should be performed bytraining developers within the Directorate of TrainingDevelopment at an Army school, prior to and during the earlyphases of concept formulation for the weapons system. The thirdfunction addresses the activity of generating training strategyoptions. It should be performed by PM TRADE engineers (in thecase of embedded training, training devices, and simulators) ortraining developers (in the case of other training approaches).The fourth function is concerned with comparing the cost-effectiveness of training strategy options. It should beperformed by training developers. The third and fourth functionsshould be completed in time for the results to be incorporatedinto the Required Operational Capability (ROC) documentation(U.S. Army Materiel Command and U.S. Army Training and DoctrineCommand, 1988).

The system is described this way for three reasons. First,it follows the sequence of steps in the ideal training devicedevelopment process shown in Figure 1. Second, it is comparableto the concept formulation process employed by the Army MaterielCommand and PM TRADE in that it requires the generation andcomparison of alternative solutions. Third, many of the ARItraining development tools tend to apply to more than one step inthe ideal process, making it difficult to organize the model bytraining development tools.

Each function is described in terms of its overall goal, itssub-functions and outputs, inputs required, benefits, anddevelopmental risks.

The benefits of the model, relative to the current devicedesign process, are listed below.

" It focuses the training device design process on preparingunits to perform their missions, the bottom line of Armytraining.

" It allows for comparing the cost-effectiveness of embeddedtraining with other options very early in weapon systemdevelopment.

7

IDENTIFY NEW COLLECTIVE AND INDIVIDUALARTEP TASKS TO BE TRAINED

DEVELOPMENTFUNCTION

ESTIMATE SAFETY, FIDELITY, AND TIMEFRONT END REQUIREMENTS FOR TRAINING TASKSANALYSISFUNCTION

IDENTIFY FEASIBLE AND PRACTICALGENERATION TRAINING DEVICE STRATEGIES

FUNCTION

COMPARE COST-EFFECTIVENESS OFTRAINING STRATEGY TRAINING DEVICE STRATEGIES

COMPARISONFUNCTION

Figure 2. Early Training Strategy Development ModelOverview.

8

" It integrates:

-individual skills training across duty positions

-individual skills training with collective training

-collective task training across unit missions

-collective task training across echelons.

* It identifies individual tasks with a high pay-off interms of progressive training value.

" It defines collective training targets of opportunity witha high pay-off in terms of progressive training value.

* It insures minimal duplication among training devices andproducts in terms of training objectives addressed.

" It provides the level of detail regarding fidelity andinstructional support feature requirements needed tocompare the cost-effectiveness of device alternatives.

" It insures that embedded training is used when it is themost cost-effective option for either individual orcollective training.

9

ARTEP Development Function

Goal. Figure 3 illustrates the information flow associatedwith the ARTEP Development Function. The goal of this functionis to estimate the effects of a new weapon system on unittraining plans as described within Army Training and EvaluationProgram (ARTEP) documents, ARTEP Mission Training Plan (AMTP) andDrill documents. A unit training plan includes:

" collective training objectives

" guidance for integrating individual task training withcollective training

" guidance for integrating collective training acrossechelons

" guidance for integrating collective training across unitmissions

" guidance for applying the progressive, building blockapproach to unit training.

Sub-functions/Outputs. ARTEP documents are the primarysource of information about the collective tasks to be performedby units. AMTP and Drill documents integrate the traditional"what to train" guidance of the ARTEP with "how to train"guidance (Hiller, Hardy and Meliza, 1984; TRADOC Regulation 310-2) to provide unit training plans which progress from individualskills training through a series of collective trainingexercises. The information from AMTP and Drill documentsrelevant to the design of integrated training device strategiesincludes guidance for integrating individual task training withcollective task training, collective task training acrossechelons, and collective task training across unit missions. Theoutput of the ARTEP function relevant to the design of trainingdevices includes the existing unit training plan for the unitreceiving a new weapon, modified to reflect the features of thenew weapon.

AMTP and Drill documents provide two types of informationabout how to integrate individual skills training with trainingon specific collective tasks. First, these documents identifythe individual tasks which should be mastered before training ona particular collective exercise. Second, they identifyindividual tasks which can be safely trained in the context ofparticular collective exercises without detracting fromcollective training. By virtue of the fact that AMTP and Drilldocuments address the skills of individuals across duty positionsand Military Occupational Specialties (MOS), these documents alsoprovide information about integrating training across dutypositions.

10

ARTEP MISSION TRAININGPLAN (AMTP) ASSOCIATED

WITH PREDECESSORWEAPON SYSTEM

ESTIMATE IMPACT NEW COLLECTIVE ANDOF NEW WEAPON ,1 INDIVIDUAL TASKS

ON TRAINING

-REVISIONS IN COLLECTIVETRAINING OBJECTVES

-CHANGES IN INDMDUALTASK PREREQUISITES FORCOLLECTIVE TRAINING

-CHANGES IN CONCURRENT

LOGIC BEHIND NEW EXECUTION OF INDMDUAL

WEAPON SYSTEM AND -CToa.E'TFVE TRAINING

-CHANGES IN

TRAINING VALUE OF

COLLECTIVE EXERCISES

Figure 3. ARTEP Function for Estimating the Effects of aNew Weapon on Unit Training Requirements.

11

ARTEP, AMTP and Drill documents define relationshipsbetween collective task training at one echelon and collectivetask training at the next higher echelon. For example, thesedocuments indicate squad level collective tasks which supportspecific platoon level collective tasks.

Collective training exercises begin with basic collectivetasks which warrant training in a repetitive Drill-like fashionand progress to Situational Training Exercises (STXs) which maycontain multiple collective tasks and Drills. The most complexcollective training events are called Field Training Exercises(FTXs). Each FTX contains a combination of STXs. Drills andSTXs have a high pay-off in terms of progressive training value,because each exercise addresses skills which apply to many unitmissions. The use of Drills and STXs as training vehicles helpsto integrate collective training across unit missions.

With those rare exceptions where the fielding of a newweapon is associated with major deletions and additions of thecollective tasks performed by a unit, unit training plans remainthe same as with the predecessor system. The integration ofindividual and collective training, the scope and training valueof Drills, and the scope and training value of STXS remainintact. In cases where the new weapon system is expected toenhance the capability of a unit, one would expect theenhancement to reflected by changes in selected performancestandards.

Drill and STX standards should require some revision if anew weapon is being fielded to help a unit address the threatsituation more effectively. Such revisions are generally tied tothe justification for the new weapon.

Input. The most important inputs for this function are theARTEP and AMTP documents for units using the predecessor weaponsystem. The ARTEP and AMTP documents for a unit with a nowweapon system are expected to be "cut and paste" versions of theprevious unit documents. The new documents may differ from theold in terms of revision of selected standards, deletion ofselected collective tasks, and addition of collective tasks.

Other important inputs for this function are those documentswhich describe the logic behind the decision to develop the newweapon system, because the logic is used to decide whichcollective tasks and which task standards need to be modified.The documents containing the logic behind new weapon systemsshould provide such information as a description of "how, what,when and where the system will be used on the battlefield and howit will interface with other systems" (Army Training SupportCenter, 1988). These logic documents include the BattlefieldDevelopment Plan (BDP) and the Organizational and Operational

12

(O&O) Plan for the new weapon system(s). The BDP (TRADOCRegulation 11-15, 1989) and O&O Plan (Army Training SupportCenter, 1988) are required to be developed by TRADOC very earlyin the formulation of the concept for a new weapon system.

Benefits Analysis. The benefits which accrue if theactivities of the ARTEP function are carried out in a timelymanner are listed below.

e It focuses the training device design process on preparingunits to perform their missions, the bottom line of Armytraining.

* It helps to integrate individual task training across dutypositions.

o It helps to insure that individual skills training iseffectively integrated with collective training exercises.

* It defines individual tasks with a high pay-off in termsof progressive training value (tasks which apply to manycollective tasks).

o It defines collective training exercises with a high pay-off in terms of progressive training value (exerciseswhich apply to many unit missions).

e It helps to insure that collective task training will beeffectively integrated across echelons and across unitmissions.

o It insures that standards are available for measuringcollective task performance.

Developmental Risk Analysis. The developmental risksassociated with this function are minor. A previously unreportedcooperative effort among ARI, the Ninth Infantry Division, andthe U. S. Army Infantry School resulted in the preparation ofARTEP documents for each of three experimental organizations; theMobile Assault Gun Battalion, the Light Attack Battalion, and theMotorized Infantry Battalion. These documents were prepared inless than two months in the form of "cut and paste" versions ofthe existing ARTEP document for Mechanized Infantry Battalions.These three Battalions differed from mechanized infantry units,and from one another, in terms of tactical docttine, weaponsystems and organizational structure. Information taken fromdraft doctrinal manuals regarding how each of these units wereintended to fight was used in deciding which collective tasksneeded to be deleted or added, and this information was also usedto decide which collective task standards required revision.

13

The work to create three "cut and paste" versions of newARTEP documents, described above, was accomplished a number ofyears ago. Two mo:.e recent events, implementation of theAMTP/Drill concepts and the Automated Systems Approach toTraining (ASAT), should make it easier to update ARTEP documentsto reflect new weapon systems, tactical doctrine andorganizational structure. AMTP and Drill documents facilitatethe revision process by providing a more economical descriptionof training requirements than is provided in traditional ARTEPdocuments. For example, the same collective task may becontained within many different unit missions. While thetraditional ARTEP document repeatedly described the same taskwithin each mission context, the improved ARTEP essentiallydescribes the task one time and indicates the missions to whichthe task applies. Bloedorn et al (1985) developed a concept forapplying computer technology to facilitate the preparation andrevision of AMTP and Drill documents. Science ApplicationsInternational Corporation (1988) prepared draft functionalspecifications and database requirements for a prototype ASAT andcompleted the prototype in 1990. TRADOC will be producing thenext version of ASAT in the 1991-92 time frame, and ASAT shouldbe fielded by mid-decade.

Front End Analysis Function

Goal. Figure 4 indicates the inputs and outputs of theFront End Analysis (FEA) Function. The FEA Function, when usedwith the ARTEP Function, should provide most of the informationabout specific individual and collective tasks needed to decidewhat training devices might be employed to train each task. Thegoals of the function are to prepare FEA data for new individualand collective tasks and to identify FEA data for existing tasksthat can be used in developing and comparing trainingstrategies. Figure 4 is not intended to imply a strict left-to-right sequence of activities. The process is instead iterativeand recurring. New tasks are input to MANPRINT Product 4, andnew tasks, safety requirements, and estimates of time required totrain tasks are inputs to OSBATS.

Sub-functions/Output. This function provides moreinformation about tasks to be trained than does the ARTEPFunction. Although the ARTEP Function provides information abouthow to best train collective tasks in the field using operationalequipment, additional information is needed to support decisionsabout whether it is most cost-effective to train those tasksusing operational equipment, embedded training, full missionsimulators, part task trainers, or some other method.

14

NEW COLLECTIVE AND SAFETY REQUIREMENTSINDIVIDUAL TASKS FOR TRAINING TASKS

CUE AND RESPONSEFIDELITY REQUIREMENTS

CONDUCTMANPRINT PRODUCT FRONT END4 ANALYSIS OF

NEW TASKS

ESTIMATES OF TIMEREQUIRED TOTRAIN TASKS

ADDITIONAL TRAININGOPTIMIZATON OF RELEVANT INFORMATIONSIMULATION BASED ABOUTTASKS.TRAINING SYSTEMS(OSBATS)

Figure 4. Front End Analysis Function for developing

training relevant information about tasks.

15

The information required to choose among training options isgrowing as new types of training methods and products are madeavailable. Since the Army currently has limited experience withmany training device options, such as embedded training andcollective training devices, a complete list of all factors toconsider, and therefore of all desirable FEA outputs, is notreadily available.

Cost and safety are the most important factors to considerwhen selecting among training device options. At times, they canoverride all other factors. For example, safety considerationsprohibit training aircraft emergency procedures in the actualaircraft, and cost considerations severely limit the firing ofactual air defense missiles for training purposes. Anotherimportant factor is the cue and response fidelity requirements ofeach task, that is, how similar to real life mission situationsthe training must be to provide meaningful opportunities forpractice and performance feedback. Fidelity is a complexmultidimensional factor which may, for example, encompassconsiderations of which of the six degrees of movement need to beemployed, visual resolution requirements, field of viewrequirements, and interactions between motion and visual issues.The importance of fidelity is not surprising, given thesubstantial increases in cost associated with higher levels ofvisual, motion, auditory and tactile fidelity (Meliza andLampton, in preparation).

Beyond those three factors, there is little agreement amongmembers of the training community on the relative importance ofother information about tasks. This other information includesthe time required to train the task to standard, and requirementsfor providing fe-dback. These additional elements must beconsidered to insure that training is truly effective and toenhance the efficiency of training.

Benefit Analysis. This function provides information abouttasks needed for the Training Strategy Development Function andthe Training Strategy Comparison Function to perform theirfunctions. The benefits of the FEA Function are realized inthese other two functions.

Inputs. The ARTEP Function should provide input to the FEAFunction to insure that individual task analysis is integratedwith collective task analysis. A top down approach to taskanalysis (TRADOC Regulation 350-7, 1985) requires that collectivetask analysis precede and provide input to individual taskanalysis.

The FEA Function should also make use of information aboutrelated or predecessor weapon systems, including trainingobjectives. In preparing these new training objectives and other

16

information, the FEA Function would again draw upon informationfrom related or predecessor tasks, because the new trainingobjectives and information would likely take the form of a "cutand paste" version of the old material. The sources ofinformation which might be used to gather FEA material shouldinclude Training Performance Data Center (TPDC) databases, andsimilar weapon systems to which the OSBATS and ET guidelines havepreviously been applied. The initial results of the FEA would ofcourse require updating to reflect changes in the concept for anew weapon system.

Developmental Risk Analysis. For this function to supportsubsequent functions in training device design effectively, thecomplete set of FEA outputs must be defined, and the outputs mustbe developed and readily available for use. Risks associatedwith these steps will be discussed in the following paragraphs.

The Army continues to identify new FEA requirements as itexplores the use of new training technologies and newapplications of existing training technology. For example,recent experience with the use of semi-automated forces (SAFOR),such as the computer simulation of friendly forces in thesimulated networking environment, have raised questions about thetypes of tasks which SAFOR can perform without detracting fromthe realism of exercises and without overburdening theinstructor/operators who help to control SAFOR actions. A hiddendevelopmental risk lies in the ability of subsequent functions inearly training estimation to address a growing variety of FEAinput data when developing and comparing training strategies.That is, the subsequent functions need to be open-ended.

The process by which FEA output is developed must evolve tomeet the information needs of training designers. Current FEAoutput does not satisfy all of the information requirements ofexisting decision rules for designing training devices. Aneffort to assess the availability of FEA material required tocompare the cost-effectiveness of training device options usingOSBATS found that some required information was not available ineither schools or elsewhere (Willis, Guha, Hunter, 1988). Arecent examination of the training device concept formulationprocess within PM TRADE found that some of the most criticalinformation about tasks, fidelity requirements are difficult toobtain (Meliza and Lampton, in preparation). These observationswere made even in situations where the fidelity issues wereindependent of the design features of a new weapon system (i.e.,fidelity issues which applied to the predecessor system and itstraining devices). There is a need to impress training analystswith the importance of these FEA outputs, and there is a need tofacilitate the preparation of these outputs.

17

Substantial progress has been made in developing procedureswhich make it easier to locate appropriate FEA output frompredecessor systems. MANPRINT Product 4, for example, will makeit easier to identify tasks comparable to those associated withnew weapons. OSBATS contains FEA output within its database.Increased application of OSBATS to training device design, andthe accumulation of OSBATS data in a centralized database, wouldgreatly increase the extent to which such data are readilyavailable.

Another important variable to consider in terms of qualityof FEA outputs is the level of expertise of analysts. Armyschools can use the selection process and training to increasethe level of expertise of task analysts.

TraininQ StrateQy Generation Function

Goal. The goal of this function is to generate multipletraining strategies for a new weapon system. Each strategygenerated should meet certain criteria (such as being affordable)to warrant further consideration (see Figl-i ). Multiplestrategies are necessary to allow for sub. sequent comparison ofthe strategies to select the one which is most cost-effective (ormost effective at or below a given cost).

Each training strategy defines the imi;,: s of trainingproducts and devices required to provide effective individual andcollective training on a new weapon system. This macro strategy,discussed in this section, should be distinguished from the microstrategies which describe how each training device option is tobe used to achieve specific training objectives (US Army Trainingsupport Center, 1988). The early development of a trainingstrategy involves specifying the training device options(opera ',Inal equipment, embedded training, full missionsimulator, networked simulations, and part task trainer) 6rcombination of options to be employed within each strategy. Forexample, one training strategy might call for all individual andcollective tasks to be trained using embedded training ininstitutions and units, while a second strategy might use acombination of training on operational equipment, part tasktrainers and networked simulation to address the same tasks asthe first strategy.

Sub-functions/Outputs. The purpose of this function is togenerate a group of training strategies for a new weapon system.Each strategy should address all of the tasks to be trained.

In addition to defining a mix of training products anddevices, a good training strategy should meet additionalcriteria. First, a strategy must indicate where each task is to

18

TRAINING DEVICEPOLICY

GUIDANCE FORAPPLYING TRAININGMETHODS

TECHNOLOGICALDATABASE

ROMCOST ESTIMATES_________

CS GENERATE TRAINING TRAINING STRATEGYSTRATEGY OPTIONS OPTIONS

OUTPUT FROMFEA FUNCTION -uMITED

DEVELOPMENTAL RISKS

- AFFORDABLE

OUTPUT FROMARTEP DEVELOPMENT -SPECIFIES TASKSFUNCTION TO BE TRAINED

ON EACH DEVICE

TRAINING FACILITIES -INTEGRATESDATABASES, TOE, INDMDUAL ANDAND MANPRINT COLLECTIVE TRAINING

-INTEGRATES TRAINING

DSBATS SIMULATION ACROSS MISSIONSCONFIGURATION AND ECHELONSMODULE

Figure 5. Training Strategy Generation Functionfor developing feasible and practical trainingstrategy options.

19

be trained for the first time (institution or unit). Second, astrategy should be presented in sufficient detail to indicate

which groups of individual and collective tasks are to beaddressed by each specific training device. This is necessary toinsure that all parties will understand the scope of the device.Third, the implementation of each strategy should involve minimaldevelopmental risks. No strategy should call for applications oftechnology that are impossible given the state of the art orwhich are so novel that insufficient information is available toassess the potential for success. Fourth, a training strategyshould reflect the integration of individual and collectivetraining, collective task training across unit missions, andtraining across echelons.

Benefits Analysis. The type of training strategiesdescribed above:

" insures that there will be minimal duplication amongtraining devices and products in terms of trainingobjectives addressed, because devices and products arelinked to specific training objectives;

" provides the level of detail required for accurateestimation of the cost of implementing training strategyoptions;

* increases the likelihood that all devices being consideredare technical'v feasible; and

" insures an overall efficiency of training by integratingthe various training requirements.

Inputs. The inputs to this function are used to insure thatany training strategies developed are feasible and practical.To be practical and feasible the strategies must be in compliancewith training device design policies, affordable, safe,technologically feasible, and compatible with the trainingenvironment. These strategies must integrate individual andcollective training, and they must integrate training acrossmissions and echelons. Finally, the strategies must be thoroughand specific in describing the tasks to be trained by eachdevice. The inputs required to develop strategies which meet allof these criteria are discussed below.

A major input to the training strategy development processis the current Department of the Army (DA) policy concerningembedded training as it is defined in a policy and guidanceletter dated March, 1987 and signed jointly by the Vice Chief ofStaff and the Under Secretary of the Army. This letter statesthat "an embedded training capability will be thoroughlyevaluated and considered as the preferred alternative among otherapproaches to the incorporation of training subsystems in the

20

development and follow on Product Improvement Programs of allArmy materiel systems". This policy was interpreted by Strasel,Dyer, Roth, Alderman, and Finley (1988) to mean that "ET will beincluded in all new and developing Army systems unless there arevalid and compelling reasons not to do so". The potentialtargets of opportunity for ET include collective training. Forexample, the Systems Training Plan for the Main Battle Tank (MBT)Block III states "embedded training will include one or morestand-alone modules for force on force simulation" (U. S. ArmyArmor School, 1989).

Training device policies may apply to many different weaponsystems, or they may apply to a single weapon system (Hinton etal., in preparation). DA guidance to consider ET as thepreferred alternative among other approaches is an example of apolicy which influences virtually all efforts to develop trainingdevices for new weapon systems. Policy may also be specific to agiven weapon system or family of weapon systems. For example, apolicy might state that the training devices must be modular andreconfigurable to allow the same basic device to be employed foreach member of a family of weapon systems.

Guidance is also required about how to apply specifictraining methods. For example, Volume 2 of the ET trainingdevelopment guidelines (Strassel et al., 1988) contains rules fordeciding which tasks and subsystems of a weapon system areappropriate to be addressed by ET. The ET guidance considers suchissues as the likelihood that ET can be provided withoutinterfering with the weapon system's operational capabilities.These rules are currently being expanded to provide more detailedguidance, in the form of logic flowcharts and accompanyingexplanations, for making early ET decisions (Witmer and Knerr, inpreparation).

The application of simulation networking to combined armstraining has added a new set of policies to be considered whendeveloping training strategies. For the simulators which supporttraining on weapon system to be networked with simulators forother weapon systems in combined arms training exercises, thesesimulators must be designed to be interoperable. A DOD levelstandard to support interoperability in the simulation networkingenvironment is in a late stage of development (McDonald, Pinon,Glasgow, and Lanisas, 1990).

It is important to identify problems in the technologicalfeasibility of device options as early as possible. Theidentification of such problems is part of the job of PM TRADEengineers, and PM TRADE is responsible for maintaining atechnological database to support the identification proc ss(U.S. Army Materiel Command and U.S. Army Training and DoctrineCommand, 1988).

21

Information is required which can be used to estimate thecost of implementing the various strategies. Given the fact thatthese estimates are to be made early in the formulation oftraining strategies for new weapon systems, these estimates areunlikely to be based on the thorough and rigorous procedures usedby PM TRADE engineers in preparing a Baseline Cost Estimate(BCE). Instead, these estimates are more likely to take the formof the rough order of magnitude (ROM) estimates developed bythese engineers (PM TRADE, 1986).

The output from the FEA Function should also provide inputto the development of training strategies. The FEA outputdescribes the individual tasks to be trained and providestraining-relevant information about both the collective andindividual tasks to be trained. This training-relevantinformation includes safety considerations, estimates of the timerequired to train each task, and other information needed todecide which training device options can be applied appropriatelyto each task.

The output from the ARTEP Development Function also providesinput to the design of training strategies. One of the mostimportant inputs to training strategy generation are the unittraining plans from the ARTEP. These plans provide theinformation needed to develop strategies which integrateindividual and collective training, and training across unitmissions and echelons. Such information is important indeveloping strategies in which individual skills trainingsupports rather than disrupts collective training. The ARTEPalso describes the collective tasks to be trained, andis describes collective training exercises (Drills and STXs).In some ways the collective training exercises provide a moreaccurate description of training requirements than do meredescriptions of collective tasks, because the exercises reflectprevious consideration of training-relevant information aboutcollective tasks. For example, the STX concept reflects the factthat certain collective tasks cannot be trained and performedseparate from other collective tasks.

The job of generating a training strategy also requiresinformation about the trainee population and the environment orenvironments in which training is to be conducted. Suchinformation can be obtained from Tables of Organization andEquipment (TOEs), MANPRINT analyses, and the DoD TrainingPerformance Data Center's database of DoD Installation Ranges andTargets. The number of soldiers to the trained, the timeavailable to conduct training, and the resources available tosupport training are important variables in developing trainingstrategies.

22

The OSBATS offers potential as a tool in generatingmultiple training strategies for a new weapon system byconsidering such factors as safety issues concerning training onoperational equipment, overlaps among tasks in terms of thestimuli necessary to provide meaningful opportunities forpractice (fidelity requirements), and device development cost.The Simulation Configuration Function of the OSBATS containsdecision rules and databases which help to categorize individualtasks into clusters based upon similarity of cue and responsefidelity requirements and safety concerns. The outcome of theapplication of the OSBATS Simulation Configuration Module is inthe form of a macro training strategy which indicates the tasksto be trained on each of three training device options(operational equipment, full mission simulator, or part missionsimulator. In addition, the Simulation Configuration Model givesthe training designer the capability to develop variations of atraining strategy by, for example, varying the importance of costas a strategy determining variable. This feature allows thetraining designer to generate multiple training strategies interms of the tasks assigned to each training device.

Developmental Risk Analysis. The process of defining atraining strategy generation function represents a tremendouschallenge. Problems to be faced include: (a) the lack of auniversally accepted process for generating a training strategy;(b) the need to generate an initial training strategy very earlyin the design of a new weapon system; and (c) the requirement toconsider unit training in detail. These problems are discussedbelow.

The overriding consideration is that a universally acceptedprocess for generating training strategies does not exist. Thisis illustrated by the fact that at least two reports regardingtraining strategies have been prepared recently (Army TrainingSupport Center, 1988; Perceptronics, 1988). These reports differin terms of decision rules and information required to applyrules. The lack of a universally accepted process for generatingtraining strategies is due, in part, to the fact that the Armyhas yet to learn how to best employ many of the "newer" trainingoptions, such as embedded training.

The TRASER model provides an overall framework for trainingstrategy decisions but lacks detailed procedures for integratingexisting tools. OSBATS and ET guidelines require integration toinsure that the full range of device options can be definedwithin a single training strategy. This integration might takethe form of applying OSBATS modules to embedded trainingcandidates (e.g., to create strategies involving different mixesof training on operational equipment, embedded full missionsimulators and embedded part mission trainers) or expanding theOSBATS to include embedded training as an option. Developmentalrisk is considered low to moderate.

23

The first problem is compounded by a second. The earlygeneration of a training strategy is dependent upon adequate FEAinput. Although much of this FEA input would be derived fromdata on comparable systems, adequate FEA data on comparablesystems are often not available. This problem was discussedunder the description of the FEA Function.

The third and most difficult problem is that the unittraining environment is much more complex than the institutionaltraining environment, and training environments vary greatlyamong units. Certain of the variables which need to beconsidered when developing a training product for use in units(i.e., personnel turbulence, personnel shortages, trainerexperience and garrison/administrative distractions and thestrategies used by units to address these problems) wereidentified in the course of a study of training detractorsconducted within five Army divisions (Funk, Johnson, Batzer,Gambell, Vandecaveye and Hiller, 1980). The unit trainingenvironment has since become more complex due to the advent ofArmy Combat Training Centers (CTCs), such as the NationalTraining Center (NTC) at Fort Irwin, California. CTCs are anextension of the unit training environment which provide unitswith the opportunity to conduct leader and collective trainingwhich often cannot be accomplished at home station due toresource constraints. CTCs have expanded the unit trainingenvironment, and, at the same time, CTCs have become a majorvariable influencing the unit training environment at home-station (Fobes and Meliza, 1989).

Guidelines and procedures need to be developed forconsidering the unit training variables listed above whendeveloping unit training strategies. These procedures andguidelines should also consider two unique aspects of collectivetraining which are expected to influence training deviceselection and design. First, many of the cues which initiate andguide performance are produced by other participants in thecollective task. It is therefore more difficult to control thesituational cues required for effective training, and it is moredifficult to provide feedback to training participants. Second,the standards for performing higher level collective tasks mustafford more latitude than those for an individual task toaccommodate the fact that alternative task steps might beperformed when executing a collective task. In addition,variations of the exact situation under which a collective taskis performed also influences the range of responses which mightbe considered acceptable.

The Army also requires decision rules and databases for usein deciding when it is appropriate to use simulation networking(SIMNET) as a training device option. SIMNET-Training (SIMNET-T) represents a prototype or test-bed for future collective andcombined arms training devices. Drucker, Campshure and

24

Campbell (1988) and Burnside (in preparation) have analyzedSIMNET-T to identify specific tasks and standards which can betrained effectively in SIMNET without modifying SIMNET hardwareand software. This information can be directly applied to thedevelopment of training strategies which encompass collective andcombined arms training. Further, SIMNET-T may serve as a tool inanswering additional questions relevant to collective andcombined arms training.

The number of distinct training strategies which might begenerated for a weapon system might be excessively large, butselecting training which are feasible and practical helps toreduce the number of options. Further the current problem withthe number of training strategies is that there are far too fewstrategies . Quite often only two strategies are considered withone of these being training on operational equipment (Meliza andLampton, in preparation).

25

Training Strategy Comparison Function

Goal. This function (Figure 6) attempts to address twogoals. The first goal is to collect additional information foruse in comparing trainng strategy options. This approximatesthe Trade-Off Determination (TOD) process performed by PM TRADEengineers (PM TRADE SOP 66, 1986). The second goal is to analyzethe information collected to identify the most cost-effectivestrategy, or to identify which strategy is most effective at orbelow a given price. This second goal approximates the Trade-off Analysis (TOA) performed by schools to select device optionsfor development.

The second goal might be considered to be too ambitious,given the inadequate information available about tasks to betrained early in the concept formulation process for a new weaponsystem. A more conservative view of the goal of early comparisonof device options has been presented by Sticha, Blacksten,Morrison, and Cross (1988). These authors suggest that it iscritical for the early comparison of device options to identifybad device options and less critical that the comparison identifythe best device option. This conservative approach was appliedin the Training Strategy Generation Function of the currentmodel. In the Training Strategy Generation Function, then, theemphasis is on the development of feasible training deviceoptions. In the Training Strategy Comparison Function, theemphasis is on the identification of the "best" option.

Sub-functions/Outputs. This function must serve as a toolfor analyzing all of the information obtained about trainingstrategy options to make sure that the option selected is cost-effective and affordable. An ideal feature of this functionwould be to provide training designers with the capability toassess rapidly the effects of varying the design features of adevice on its relative cost-effectiveness. The cost of a deviceor mix of devices can be reduced by designing it to increase theefficiency of training, thereby reducing the number of devicesrequired. For example, instructional support features can beincluded in the design of a device to increase its efficiency byreducing administrative "down time".

The information which must be derived from this function tosupport fully the comparison of training device alternatives islisted below. This list reflects both the ideal goal ofselecting the best option and the more practical goal of avoidingthe more inferior options.

" An estimate of the life cycle cost of implementingparticular training strategies

" Comparisons of the relative costs of implementing strategyoptions

26

TRAINING STRATEGYOPTIONS

COST-EFFECTIVENESSOF STRATEGY OPTIONS

COST DATA /

FOR FIDELITYOPTIONS COPR

COMPARECOST-EFFECTIVENESSOF COMBINATIONS OFTRAINING DEVICEOPTIONS

COST DATAFOR INSTRUCTIONALFEATURES

MOST EFFECTIVEDEVICE AT AGIVEN COST

ADDITIONAL DATAREGARDING TASKSOR DEVICEOPTIONS

Figure 6. Training Strategy Comparison Functionfor estimating relative cost-effectiveness of trainingstrategies.

27

" Changes in the cost of implementing a training strategyresulting from shifting selected tasks from one trainingdevice to another

* Changes in the cost of implementing a training strategyresulting from the addition or deletion of instructionalsupport features.

Benefit Analysis. The Training Strategy Comparison Functionwill contain data bases and decision rules which can be used to:

" compare cost estimates of alternative trainingstrategies which include major training device options;

" provide reasonably valid estimates of the costof training on a new system for incorporationwithin budgeting requests; and

* identify strategirs which tend to be the leastcost-effectiv-

Inputs. In ori-- to price various training strategies andcompare their effeutiveness, this function must be capable ofaccepting from cther sources, storing, or developing data on theamount of time and resources required to meet training objectivesusing various training devices and products. In addition, thisfunction must be able to access or store information about theenvironments in which training products and devices are expectedto be used. Information such as number of students and courselength is required to estimate the number of training devicesrequired.

Data from the FEA Function would also be used by theTraining Strategy Comparison Function. Data currently producedby the FEA process include information such as estimates offrequency with which soldiers must be trained on specificindividual and collective tasks to sustain performance. Suchinformation is critical in determining annual resourcerequirements for training.

Finally, this function again requires life cycle costestimates of training device options, fidelity levels, andinstructional support features.

Developmental Risk Analysis. It is within the TrainingStrategy Comparison Function that the interface with existingtraining development aids is greatest. OSBATS, in particular,performs many (but not all) of the activities required. Thefollowing activities, however, are not performed by any existingaids and would need to be developed.

28

" A tool is required to compare the cost-effectiveness ofembedded training with other options. This shouldinclude both fully embedded training (all traininghardware and software are a relatively permanent part ofthe operational weapon system) and strap-on embeddedtraining (training hardware and software temporarilyinterface with the operational weapon system).

* A tool for estimating the cost of collective trainingmust be developed. A substantial portion of informationrequirements unique to collective training which areneeded to accomplish this expansion are in the form ofoutputs of the ARTEP Development Function (such asinformation regarding the integration of specificindividual and collective tasks). The Standard ArmyTraining System (SATS), wnin developed, may at leastpartially satisfy this need (U.S. Army Training andDoctrine Command, 1989)

" Tools for training device selection and resourceallocation must be developed which include the unit, aswell as the institutional, training environment. Thisexpansion is necessary to insure that accurate estimatesof the number of training devices required will bedeveloped. For example, the Resource Allocation portionof the OSBATS constructs schedules based upon trainingmethods and resources available within an institution,and these resources and methods differ between schoolsand the c-perational environment. Further, trainingconstraints and practices vary among operational units inways which influence both the need for training devicesand products and the ability to use these devices. Atool is required which incorporates unit trainingvariables which influence the relative cost-effectivenessof device )ptions.

" Techniques for developing life cycle cost estimatesinstead of, or in addition to, developmental costs arerequired.

29

Work Required to Implement Model

Implementation of the Early Training Strategy DevelopmentModel requires that research first be conducted to answer thequestions listed below. As indicated in parentheses, most ofthese issues are already being addressed by various ARI FieldUnits.

" What are the unit training environment variables thatinfluence individual and collective training in units, andwhat are the effects of these variables? (The Presidio ofMonterey and Fort Benning Field Units are addressing theseissues in the context of efforts to examine therelationships between home-station and CTC training.)

" What general approach should be taken to integrate ETguidelines with the OSBATS concept?

" What device options and instructional supportfeatures support the integration of individualand collective training?

" What device options and instructional support featuresaid in the conduct of combined arms training? (This issueis being addressed by the Fort Knox and PM TRADE FieldUnits.)

" What rules should guide the application of semi-automatedforces in the design of collective training devices?(The Fort Knox Field Unit may address this issue in futureyears.)

" What are the fidelity requirements for conductingcollective and combined arms training in a simulationnetworking environment, and which collective and combinedarms tasks can be trained cost-effectively in a simulationnetworking environment? (The Fort Knox and PM TRADEField Units are addressing these issues.)

" What general approaches should be taken to integratedecision rules and databases regarding additionaltraining device options with the OSBATS? (This issue isbeing addressed by the ARI PM TRADE Field Unit)

31

Summary

A truly top-down approach to defining a training systemarchitecture must be directed towards supporting collectiveperformance in the field. The scope of the architecturemust reflect a consideration of practical concerns suchas the large number of individual and collective trainingrequirements to be addressed, integration of training devicedesign with weapon system design, integration of individual andcollective training, the compatibility of training devices andproducts with the complex training environment, and costconstraints.

Consideration of alternative training devices and devicedesign features early in the development of a new weapon systemshould insure that affordable and cost-effective training deviceoptions are selected. No mechanism exists to support suchconsiderations early or otherwise, due to gaps in availabledecision rules and databases. However, the ASAT, OSBATS, ETguidelines, and MANPRINT Product Four provide an importantbeginning to the development of the needed mechanism. Refinementof the concept described in this report will help to make earlytraining strategy development a reality.

33

REFERENCES

Armstrong, J.N. & Deaver, R.E. (1990). "Training in the 21stCentury...The Army's Long Range Training Plan" ArmyResearch, Development and Acauisition Bulletin. January-February, 6-7.

Bloedorn, G. W., Crooks, W. H., Merrill, M.D., Saal, H.J.,Meliza, L.L., & Kahn, 0.1. (1985). Concept Study of theComputer-Aided ARTEP Production System (CAPS) (ARI ResearchReport 1403). Alexandria, VA: U. S. Army Research Institutefor the Behavioral and Social Sciences. (AD A170 951)

Burnside, B., (1990). Assessing the Capabilities of TrainingSimulations: A Method and Simulation Networking (SIMNET)Application (ARI Research Report 1565). Alexandria, VA:U.S. Army Research Institute for the Behavioral and SocialSciences. (AD A226 354)

Christie, J. D. (1987, February) Army Science Board Ad HocSubgroup on Army Analysis [Letter to the DeputyUndersecretary of the Army, Operations Research).

Department of the Army (1987, March). Policy and Guidance Letter,Subject: Embedded Training. Washington, DC.: Author.

Ditzian, J. L., Roth, J. T., & Johnston, E. (1987) DesignSpecification for a MANPRINT Training CharacteristicsEstimation Aid (TCEA). Butler, PA: Applied ScienceAssociates, Inc.

Drucker, E. H., Campshure D. A., & Campbell, R. C., (1988).An Analysis of Tank Platoon Operations and Their Simulationon SIMNET (FR-PRD-88-21). Alexandria, VA: Human ResourcesResearch Organization.

Dynamics Research Corporation (1984). A Study of Field Practicesin Manpower, Personnel, and Training Documentation for theLCSMM (E-9168U). Wilmington, Author.

Finley, D. L., Alderman, I. N., Peckham, D. S., & Strasel, H.C. (1988). Implementing Embedded Training (ET): Volume 1 of10: Overview (Research Product 88-12). Alexandria, VA: U.S. Army Research Institute for the Behavioral and SocialSciences. (AD A201 401)

Fobes, J. L. & Meliza, L. L. (1989) Integrating National TrainingCenter Feedback into Home Station Training Management. ARIResearch Report (Research Report 1523) Alexandria, VA: U.S. Army Research Institute for the Behavioral and SocialSciences. (AD A213 436)

35

Funk, S., Johnson, C., Batzer, E., Gambell, T., Vandecaveye,G. and Hiller, J. (1980). Training Detractors in FORSCOMDivisions and How They are Handled (Research Report 1278).Alexandria, VA: U. S. Army Research Institute for theBehavioral and Social Sciences. (AD A099 188)

Heeringa, L.M., Baum, D.R., Holman, G.L. & Peio, K.J. (1982).An Analysis of Training Device Recquirements DocumentationProcedures (Final Report). Minneapolis, MN: HoneywellSystems and Research Center.

Hiller, J.H., Hardy, G.D. & Meliza, L.L. (1984). Guideline forDesigning Drill Training Packages (ARI Research Product84-15). Alexandria, VA: U. S. Army Research Institute forthe Behavioral and Social Sciences. (AD A168 579)

Hinton, W.M., Braby, R., Feuge, R.L., Stults, A.H., Evans, S.M.,Gibson, M.R., & Zaldo, W.T. (1990). A Design Architecturefor an Integrated Training System Decision Support System(ARI Research Report 1566). Alexandria, VA: U.S. ArmyResearch Institute for the Behavioral and Social Sciences.(AD A225 084)

Hofer, R., Ozkaptan, H. and Kincaid, J.P. (1987) "TrainingSystems R&D Program: Progress and Challenges",Interservice/Industry Training Systems Conference.

Mannle, T.E.,Jr., Guptill, R.V., & Risser, D.T. (1985).HARDMAN Comparability Analysis Methodology Guide,Volume I, Managers Guide (ARI Research Product 85-19).Alexandria, VA: U.S. Army Research Institute for theBehavioral and Social Sciences. (AD A156 787)

McDonald, L.B., Pinon, C., Glasgow, R., Danisas, K. (1990).Military Standard (Draft): Protocol Data Units forDistributed Interactive Simulation (IST-PD-90-2).Orlando, FL: Institute for Simulation and Training.

Meliza, L., & Lampton, D. (1991). Engineering Functions inFormulating Training Device Concepts (ARI Research Report1594). Alexandria, VA: U.S. Army Research Institute forthe Behavioral and Social Sciences.

Perceptronics, Incorporated (1988). Traininq system andSubsystem Optimization Model Draft Report. Woodland Hills,CA: Author.

36

Peden, I.C., Barth, D.S., Bonder, S., Caro, P., Fried, D.L.,Jones, E.R., LaBerge, W.B., Morrison, L.W., O'Neal, R.D., Pauly, J.W., Pettigrew, K.D., Rathjen, K.R., Simmons,C.J., Welch, J.J., Williges, R.C., Zarafonetis, C.J.D.(1985). Final Report of the 1985 Summer Study on Trainingand Training Technology - Application for AirLand Battleand Future Concepts. Washington, D.C.: Army Science Board,Assistant Secretary of the Army, Research, Development andAcquisition.

Roth, J.T., Warm, R.E., Peters, J., O'Brien, L., Hawley, J.K.,Pence, E.C., Robinson, R.E., Masterson, S., & Criswell,E.L. (1989). MANPRINT Methods Monograph: Aiding theDevelopment of Training Constraints (ARI Technical Report850). Alexandria, VA: U.S. Army Research Institute forthe Behavioral and Social Sciences. (AD A215 636)

Roth, J. T., Warm, R. E., Peters, J., Masterson, S. &Criswell, E.(1987). Concept for a MANPRINT TrainingCharacteristics Estimation Aid (Product Four). Butler,PA: Applied Science Associates.

Science Applications International Corporation (1988). AutomatedSystems Approach to Training Functional Description. SanDiego, CA: Author.

Sticha, P. (1988). Research and Development on theCharacterization of Simulation-Based Training Systems:Project Executive Summary (HumRRO Final Report FR-PRD-88-28). Alexandria, VA: Human Resources Research Organization.

Sticha, P., Blacksten, H. R., Buede, D.M., Singer, M.J.,Gilligan, E.L., Mumaw, R.J. & Morrison, J. E. (1988).Optimization of Simulation-Based Training Systems: ModelDescription, Implementation and Evaluation (HumRRO FinalReport FR-PRD-88-26). Alexandria, VA: Human ResourcesResearch Organization.

Sticha, P. J., Singer, M. J., Blacksten, H. R., Morrison, J. E.,and Cross, K.D. (1988). Research and Methods for SimulationDesign: State of the Art (HumRRO Final Report FR-PRD-88-27). Alexandria, VA: Human Resources Research Organization.

Strasel, H.C., Dyer, F.N., Roth, J.T., Alderman, I.N. & Finley,D.L. (1988). Implementing Embedded Training (ET): Volume2 of 10: Embedded Training as a System Alternative (ResearchProduct 88-22). Alexandria, VA: U.S. Army Research Institutefor the Behavioral and Social Sciences. (AD A204 836)

U.S. Army Armor School (1989). Systems Training Plan for the MainBattle Tank (MBT) Block III. Fort Knox, KY: Author.

37

U.S. Army Armor School (1990). Training Device Requirement (TDR)for the Close Combat Tactical Trainer (CCTT). Fort Knox, KY:Author.

U.S. Army Materiel Command and U.S. Army Training and DoctrineCommand (1988). Concept Formulation Process (CFP) Memorandumof Instruction (MOI). Alexandria, VA and Fort Monroe, VA:Authors.

U.S. Army Project Manager for Training Devices (1986). ProjectManagement and Execution (Standard Operation Procedure 66).Orlando, FL: Author.

U.S. Army Training and Doctrine Command (1982). Individual andCollective Training Plan for Developing Systems; Policy andProcedures (TRADOC Regulation 351-9). Fort Monroe, VA:Author.

U.S. Army Training and Doctrine Command (1984). Design,Development, Preparation, and Management of ARTEP Documents:Mission Training Plans (MTPs) and Drill Books (TRADOCReg 310-2). Fort Monroe, VA: Author.

U.S. Army Training and Doctrine Command (1985). Systems Approachto Training (TRADOC Regulation 350-7). Fort Monroe, VA:Author.

U.S. Army Training and Doctrine Command (1989). Systems TrainingDevelopment (Draft TRADOC Regulation 351-9). Fort Monroe,VA: Author.

U.S. Army Training and Doctrine Command (1989). Concept BasedReguirements System (TRADOC Regulation 11-15). Fort Monroe,VA: Author.

U.S. Army Training and Doctrine Command (1989). Standard ArmyTraining System (SATS) Menu/Screen Dictionary. Fort Monroe,VA: Author.

U.S. Army Training Support Center (1988). Training Developer'sProcedural Guide: Training Device Strategies. Fort Eustis,VA: Author.

Willis, R.P., Guha, P., Hunter, D. (1988). The Optimization ofSimulation-Based Training: Model Data Collection andUtilization. Orlando, FL: Eagle Technology, Inc.

Witmer, B.G. and Knerr, B.W. (1991). A Guide for Early EmbeddedTraining Decisions (ARI Research Product 91-14).Alexandria, VA: U.S. Army Research Institute for theBehavioral and Social Sciences. (AD A239 669)

38