23902012025212S
THE PROPOSED RESEARCH PROJECT WILL (I) STUDY THE FLEET TRAVERSING SCHEME DESIGN PROBLEM, (II) DEVELOP A GP MODEL TO INTEGRATE THREE ESSENTIAL GOALS, (III) ANALYZE DATA ENVELOPMENT METHODS TO MEASURE EFFICIENCY, CONSTRUCT A SPREADSHEET TO ACCOMMODATE GP AND DEA METHODS SIMULTANEOUSLY, AND APPLY A CASE STUDY TO DEMONSTRATE THE APPLICABILITY OF THE PROPOSED PROCEDURE. THE SECTIONS THAT FOLLOW PRESENT THE DETAILS OF THE MAJOR TASKS TO BE PERFORMED AND THE PROPOSED TIMELINE. RESEARCH TASK 1 Â LITERATURE REVIEW THIS TASK IS TO IDENTIFY PRIOR WORK SUITABLE FOR THIS RESEARCH THROUGH AN EXPANDED REVIEW OF RELEVANT LITERATURE. THE STUDY WILL HELP THE RESEARCH TEAM IDENTIFY SOURCES WHERE MODEL INPUT DATA CAN BE OBTAINED. ALSO, THE RESEARCH TEAM WILL SURVEY CURRENT PRACTICES TO MANAGE THE RISK OF ROUTE DISRUPTIONS. THE PURPOSE OF THE SURVEY IS TO BETTER UNDERSTAND THE MANY DIFFERENT FACTORS AFFECTING FLEET TRAVERSING AND MAKE SURE THE RESEARCH RESULTS ARE THEORETICALLY AND PRACTICALLY SOUND. RESEARCH TASK 2 Â DATA COLLECTION ACCURATE AND DETAILED INPUT DATA ARE CRITICAL FOR MODELING A GOAL PROGRAMMING MODEL TO INTEGRATE THREE ESSENTIAL GOALS UNDER THE RISK OF ROUTE DISRUPTIONS. THE RESEARCH TEAM WILL IDENTIFY A COMPREHENSIVE LIST OF ESSENTIAL COMPONENTS AND OPERATIONAL PARAMETERS AFFECTING THESE THREE GOALS. RESEARCH TASK 3 Â DEVELOPMENT OF OPTIMIZATION MODEL BASED ON THE EVALUATION OF EXISTING MODELS REVIEWED IN TASK 1, THE RESEARCH TEAM WILL DEVELOP A PROCEDURE FOR HOW TO FORMULATE A DESIGN PROBLEM OF THE FLEET TRAVERSING SCHEME (FTS) UNDER SINGLE-AND MULTI-OBJECTIVE OPTIMIZATION THROUGH GOAL PROGRAMMING (GP). THE MODEL WILL BE DEVELOPED FOR EACH GOAL TO CONSIDER BETTER THE UNCERTAINTIES, WHICH INCLUDES THE UNCERTAINTIES DUE TO DISRUPTIONS. THIS APPROACH IS THE ROBUST OPTIMIZATION APPROACH WHICH IS USEFUL IN THE ABSENCE OF SOUND PROBABILITY DISTRIBUTIONS. THE ROBUST OPTIMIZATION MODEL IS A LARGE-SCALE MIXED-INTEGER LINEAR/NONLINEAR PROGRAM (MILNP) WITH MANY BINARY VARIABLES ALONG WITH NUMEROUS CONSTRAINTS CAPTURING NETWORK STRUCTURE AND SYSTEM TACTICAL/OPERATIONAL REQUIREMENTS. DOCUSIGN ENVELOPE ID: 95EFBB21-3399-42C8-9F06-5500ECE941BD VIRTUAL PROTOTYPING OF AUTONOMY-ENABLED GROUND SYSTEMS EFFICIENCY-BASED FLEET TRAVERSING SCHEME DESIGN JUDITH MWAKALONGE 4 RESEARCH TASK 4 Â CONSTRUCT SPREADSHEET MODELS THE OPTIMIZATION MODELS CAN BE SOLVED BY A VARIETY OF OPTIMIZATION SOFTWARE PACKAGES, SUCH AS LINDO, LINGO, OR GAMS. HOWEVER, CODING THE DEVELOPED MILNP MODEL USING THESE TOOLS MAY NOT BE EASY SINCE SO MANY DECISION VARIABLES AND CONSTRAINTS ARE INVOLVED. MANY RESEARCHERS AND PRACTITIONERS ARE RECENTLY PAYING SIGNIFICANT ATTENTION TO MICROSOFT EXCEL SPREADSHEET-BASED OPTIMIZATION MODELING BECAUSE OF ITS NON-ALGEBRAIC APPROACH. SEVERAL POWERFUL SOFTWARE PACKAGES BASED ON THE EXCEL SPREADSHEET MODEL, SUCH AS FRONTLINE SOLVER, WHATÂ S BEST!, CPLEX, ETC., MAKE EXCEL SPREADSHEET-BASED MODELING ATTRACTIVE. IN THIS RESEARCH PROJECT, FRONTLINE SOLVER WILL BE USED TO SOLVE THE PROPOSED MILNP MODELS. RESEARCH TASK 5 Â DEVELOPMENT AND APPLICATIONS OF DATA ENVELOPMENT ANALYSIS METHOD THE RESEARCH TEAM WILL MODIFY THE MATHEMATICAL FORMULATION OF THE DATA ENVELOPMENT ANALYSIS (DEA) TO APPLY TO THE ALTERNATIVE OPTIONS GENERATED BY THE GP MODEL FORMULATED IN TASK 3, TO ASSESS THE ALTERNATIVE OPTIONS FOR IDENTIFYING THE MOST EFFICIENT AND RESILIENT OPTIONS FROM ALL. DEA-FRONTIER WILL BE USED FOR THE APPLICATION OF DEA. IF NECESSARY, THE RESEARCH TEAM WILL DEVELOP ITS COMPUTER PROGRAM FOR THE APPLICATION OF DEA. RESEARCH TASK 6 Â CASE STUDY SEVERAL CASE STUDIES WILL BE DEVELOPED TO DEMONSTRATE THE APPLICABILITY OF THE DESIGN AND BENCHMARKING FRAMEWORK FOR THE FTM. THE DATA COLLECTED IN TASK 2 WILL BE USED AS THE INPUT FOR THE CASE STUDY, AND ADDITIONAL DATA MAY NEED TO BE COLLECTED. THE RESEARCH TEAM FIRST WILL GENERATE A SET OF ALTERNATIVE SOLUTIONS BY SOLVING THE SPREADSHEET MODEL FOR THE GP MODELS BY USING THE SOLVER SOFTWARE PACKAGES.
South Carolina State University
Cooperative Agreement W56HZV2120001
$293.9k 3/1/22 26782012016334S
THE MAIN OBJECTIVE IS TO ADDRESS SOME OF THE EXISTING LIMITATIONS OF OFF-ROAD ENVIRONMENT DETECTION FOR AN AUTONOMY-ENABLED GROUND VEHICLE. SPECIFICALLY, THE PRIMARY GOALS ARE TO: 1. CONDUCT EXPERIMENTS ON OBSTACLE DETECTION FOR OFF-ROAD ENVIRONMENTS. 2. CREATE AND PUBLISH LABELED DATASETS FOR OFF-ROAD ENVIRONMENTS TO ENABLE FURTHER RESEARCHERS TO CONDUCT MORE EXPERIMENTS ON OFF-ROAD DETECTION SYSTEMS TO IMPROVE AGV OFF-ROAD NAVIGATION. THIS PROJECT'S EXPECTED MAIN CONTRIBUTION IS TO PRODUCE LABELED DATASETS FOR OFF-ROAD DETECTION, WHICH COULD SIGNIFICANTLY IMPROVE MILITARY AGV FLEET OFF-ROAD ROUTE TRAVERSING MISSIONS.
South Carolina State University
Cooperative Agreement W56HZV2120001
$439.9k 12/31/23 26702012016671S
OUR RESEARCH APPROACH FOCUSES ON THE DEVELOPMENT OF A MULTIMODAL HDT FRAMEWORK THAT IS AT THE INTERSECTION OF HUMAN COGNITIVE LEARNING, AUTONOMOUS SYSTEMS, AND IMMERSIVE SIMULATIONS OF COMBAT SCENARIOS WITH THE GOAL OF SUPPORTING THE NEXT GENERATION COMBAT VEHICLE. THE HDT FRAMEWORK WILL PROVIDE A TESTBED FOR THE NEXT GENERATION SYNTHETIC CREW MEMBERS AND THE DYNAMICS OF THEIR INTERACTIONS WITH HUMAN TEAMMATES TO OPTIMIZE THEIR PERFORMANCE WHILE REDUCING HUMAN CREW COGNITIVE LOAD IN NGCV. THIS FRAMEWORK WILL BE DESIGNED IN A MODULAR OPEN SYSTEMS ARCHITECTURE APPROACH TO ALLOW FOR ITERATIVE REFINEMENT OF THE HDT COMPONENTS, INTEGRATION WITH NGCV SIMULATIONS AND LONG-TERM SYSTEM SUSTAINABILITY. THE HDT FRAMEWORK WILL CAPTURE REAL-TIME MULTIMODAL TRACE DATA (E.G., CONCURRENT VERBALIZATIONS, EYE MOVEMENTS, PHYSIOLOGICAL SENSORS, GESTURES) FROM A HUMAN CREW MEMBER, THE VEHICLE AND COMBAT DATA AND FEED THAT DATA TO THE DIGITAL TWIN TOWARDS THE DEVELOPMENT OF ADVANCED AI METHODS THAT WILL CREATE ITS CAPABILITIES OF SENSING, LEARNING, REASONING, AND REFLECTING ON THEIR ACTIONS AND INTERVENTIONS BASED ON THEIR INTERACTIONS WITH THEIR HUMAN COUNTERPARTS. THE HDT FRAMEWORK WILL BE BUILT TO SUPPORT THE DEVELOPMENT OF EVOLVING INTELLIGENCE (GENERATIVE AND EVOLUTIONARY) MODELS BASED ON THE DIGITAL TWIN'S ABILITY TO LEARN NEW STATES FROM SOLDIERSÂ COGNITIVE, AFFECTIVE METACOGNITIVE, MOTIVATIONAL, AND SOCIAL PROCESSES CAPTURED AT DIFFERENT TIME SAMPLING RATES USING VARIOUS DATA SOURCES (E.G., CONCURRENT VERBALIZATIONS), DEVICES (E.G., EYE TRACKERS, 3D CAMERAS), AND UNDER DIFFERENT CONTEXTS (E.G., LAB STUDY, IMMERSIVE BATTLEFIELD SIMULATION, REAL-WORLD BATTLEFIELD SCENARIO). MORE SPECIFICALLY, INDIVIDUAL DIFFERENCES, AS WELL AS OTHER SELF-REPORTED MEASURES (E.G., EXPERTISE LEVEL, WORKING MEMORY CAPACITY) WILL BE USED TO INITIALIZE THE HDTÂ S INTELLIGENCE. STILL, THEN WE WILL ENGAGE IN SEVERAL CYCLES OF MULTIMODAL MULTICHANNEL DATA COLLECTION WHERE THE SOLDIERÂ S REAL-TIME MULTIMODAL MULTICHANNEL DATA WILL BE FED Â DIRECTLYÂ AND COMMUNICATED TO THE HDT (WITH A HUMAN IN THE LOOP TO MAKE INFERENCES AND TRANSLATE THE MULTIMODAL DATA SIGNALS INTO ACTIONABLE STATES) THAT WILL BE EMBODIED IN THE HDT AS PART OF ITS INTELLIGENCE CAPABILITIES (AZEVEDO ET AL., IN PRESS; KRIEGER ET AL., 2022; MOLENAAR ET AL., 2023). WE WILL ENGAGE IN SEVERAL CYCLES OF DATA COLLECTION, ANALYSIS, TRANSLATION, EMBODIMENT, AND TESTING OF THE HDTS ACROSS SEVERAL TASKS AND BATTLEFIELD CONTEXTS TO TEST THEIR INTELLIGENCE CAPABILITIES AS A TEAM MEMBER. THE HDT FRAMEWORK WILL RESULT IN A MULTI-PLATFORM IMMERSIVE SIMULATION PLACING HUMAN SOLDIERS, THE HDT AND THE NGCV IN COMBAT SITUATIONS IN WHICH THEY NEED TO INTERACT AND DEVELOP TRUST IN EACH OTHER THROUGH THEIR TRAINING EXPERIENCES.
The University Of Central Florida Board Of Trustees
Cooperative Agreement W56HZV2120001
$4.8m 12/31/23