Post on 24-Apr-2022
Social: #ADLiFEST | WiFi: HILTON_MEETING / Password: ADLiFEST
Training Management and STE: Mapping Squad Level Competencies in
Simulation Data Through a GIFT/TLA Integration
Benjamin Goldberg, Ph.D. U.S. Army Combat Capabilities Development Command-
Soldier Center STTC
Additional Authors: Gregory Goodwin, Ph.D. & Joan Johnston, Ph.D.
Concurrent Presentation SessionBRIDGING ADL AND SIMULATION
UNCLASSIFIED
U.S. ARMY COMBAT CAPABILITIES DEVELOPMENT COMMAND –SOLDIER CENTER
Benjamin Goldberg, Ph.D.
Senior ScientistCCDC-SC, Simulation and Training Technology Center
Training Management and STE (Synthetic Training Environment):Mapping Squad Level Competencies in Simulation Data through a
GIFT/TLA Integration
DISTRIBUTION A: This is approved for public release distribution is unlimited.Collaborators: Dr. Greg Goodwin (CCDC-SC), Dr. Joan Johnston (CCDC-SC), Dr. Keith Brawner (CCDC-SC)
Kevin Owens (ARL, UT-Austin) and Kevin Gupton (ARL, UT-Austin)
UNCLASSIFIED 3
ADAPTIVE PEDAGOGY AT THE TEAM LEVELLeveraging Sports and Military Psychology to Accelerate Team Development
UNCLASSIFIED 4
WHAT IS THE GOAL OF A TEAM-AIS?To Take a Team of Experts and to Build an Expert Team!...
…by leveraging sports and military psychology to accelerate team development
UNCLASSIFIED 5
OVERARCHING TECHNOLOGY OBJECTIVE
• Leverage and integrate existing opensource technology to support an experiential training ecosystem based on X-Learning model
– Implement persistent performance tracking across competency model based on SQUAD LETHALITY and TEAM DEVELOPMENT (xAPI)
– Use ecosystem to store STE exercise data for AI and adaptive training optimization and scenario generation (Big Data)
– Maximize AAR and post-STE training efficiencies through directive Training Support Packages based on Unit Deficiencies (Recommender Engines)
UNCLASSIFIED 6
• STE produces Raw Data [TSS] (simulation events [S/SVT, SAMT, IVAS, RVCT, etc.], movement, performance, bio/physio, etc.)
• Raw [TSS] consumed by STE TMT [e.g., GIFT] and produces Metrics (ACTIVE performance states [SPM/IVAS], PASSIVE performance states [MASTR-E], Go/No-Go, TEAM states [STTC research])
• STE TMT generates xAPI (experiential statements on all data context)• xAPI routed to LRS/DataLakes for storage• LRS/DataLakes accessed by Competency and Skills System (CaSS)• CaSS updates Squad Lethality Competencies based on data assertions from
STE TMT• TMT Squad Lethality linked to ATIS/TMT/TLA• Recommender engines target Solider development opportunities and guide
and/or automate plan/prepare scenario generation activities
UNCLASSIFIED 7
STE Exercise A
STE Exercise B
Competency and Skill System - Individual
Warfighter Function (Role)
• K• S• A• s
Leadership
• K• S• A• s
Teamwork
• Adaptability• Resiliency• Communicate
Experience Capture Experience Capture
xAPI
(Mea
sure
s of P
erfo
rman
ce/E
ffect
iven
ess)
TSP Generator/Library
STE TMT Data Lake
xAPI(Measures of Perform
ance/Effectiveness)
Training Strategy Recommender Engine (Solider Development)
Role Functions(each KSA)
• Technique• Mechanics• Consistency
Role in Team Context
• Declarative• Procedural• Interdepend
Team Execute Skills
• Opp Unders• Counter Tac• SA
Competency and Skill System – Team/Unit
Performance
• TTPs• Go/No-Go• Communicate
Teamwork
• Adaptability• Cohesion• Communicate
Mission
• Type• Environment• Enemy Level• Leadership
Present
STE Exercise Generator/Recommender(Objective - T)
Team Compet(each KSA)
• Technique• Mechanics• Consistency
PerformanceGo/No-Go
• Declarative• Procedural• Interdepend
OBJ T Readiness
• Mission Obj• OWT • PMESII
Plan Prepare
Intervention 1 Intervention 2 Intervention 3 Intervention 4
UNCLASSIFIED 8
STE EXPERIENTIAL LEARNING (STEEL) PROJECT (OWENS, GUPTON & GOLDBERG)
UNCLASSIFIED 9
SQUAD ADVANCED MARKSMANSHIP TRAINER (SAM-T):BATTLE-DRILL TRAINING
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
10
EXPERIENCE CAPTURE/INTELLIGENT TUTORING
Weapon:Aim (x,y)Trigger Squeeze
BioHarness:BreathingECGPosture
Performance :Hit/MissCasualtiesRate of Fire
Feedback/Remediation
STE TMT
Expert Marksmen
Models
SAM T
BioHarnessSensor
Weapon
Performance
Eye-Tracking(Nov. 19?)
Raw DataTeam
CoordinationModel
TeamPerformance
Model
Back-Up BehaviorModels
Team/Trainee Model
Coaching Model
Assessment
Total LearningArchitecture
LMS
Learner Record StorexAPI
Competency Framework:SQD Performance Model
Recommender Engine
Universal Trainee Record
Eye Tracking:Gaze/FixationCommunication:NLPOC Input
Scenario Library
DKF Metadata Matching Metadata File
Scenario Request
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
11
SAM TWeapon
Performance
Eye Track
BioHarness
GatewayRaw
AMQDomainModule
TraineeModule
PedModule
LMS
LRS
LOG Offline
Perfo
rman
ceSt
rate
gyStrategy
Ped Request
Trai
nee
Stat
e
PedRequest
xAPI
Strategy
Raw
GENERALIZED INTELLIGENT FRAMEWORK FOR TUTORING (GIFT)
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
12
Presence Patrol
ROE Weapon Control Reporting
xAPI Data Lake
xAPIxAPI
xAPI
Soldier1 Movement
Competency
Soldier1 Shooting
Competency
Competency and Skills System (CaSS) Mapping
Soldier2 Movement
Competency
Soldier2 Shooting
Competency
SquadModel
CaSS Mapping (Unit Level)
CaSS Mapping (BTL Level)
Scenarios create these via GIFTxAPI specifications
(authored by domain experts…Represented in GIFT task model)
Training Managers create these
Training Managers create these
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
13
STE EXPERIENTIAL LEARNING (STEEL) PROJECT (OWENS, GUPTON & GOLDBERG)
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
14
ARCHITECTURE STRATEGY
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
15
Moving Forward…
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
16
• Receive proponent approval to proceed (COMPLETE)• Build Testbed Environment for Data Logging
• Modify API/GIFTadapter for SAM T data formats (COMPLETE)• Incorporate eye-tracking and video/motion capture data sources into modeling efforts
(IN PROGRESS)• Coordinate with operational training test sites identified for pilot data collection
• Fort Benning (Initial Entry Training)• Fort Bragg (Special Operational Forces)• Fort Campbell (Home Station Training)
MOVING FORWARD…
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
17
• Define and Implement a Data Strategy (WE need your help!)• Proper Front-End Analysis with Community of Interest
• Establish analysis Wish List (What questions do you want the data to answer?)
• Establish simulation translation layer via GIFT with specified measures mapped to analysis requirements
• Build xAPI statement structures around GIFT logging schemas• Build Reference Implementation of Squad Performance Model (SPM) in CaSS
• Research individual competency model to support Solider development
• Research team competency model predictive of readiness (Objective – T)• Implement Data-Driven recommender engines
• Training Strategy grounded in sports and military psychology
• Targets accelerated team-development centered on lethality and SPM attributes
MOVING FORWARD…
UNCLASSIFIED
UNCLASSIFIED//FOR OFFICIAL USE ONLY//DRAFT//PRE-DECISIONAL
18
THANK YOU FOR YOUR ATTENTION!!!