Effect of Autonomous Assistance on Crew Member Performance, Situational Awareness, and Interaction in a Closed Hatched Simulated Environment

Principal Investigator

• Despina Stavrinos, University of Alabama at Birmingham (UAB)

Faculty/Staff

• Benjamin McManu, Piyush Pawar, Andrew Underhill, Samuel Misko, James Michael Brascome, UAB
• Tulga Ersal, U. of Michigan

Students

• TBD, UAB

Government

• Victor Paul, Terry Tierney, US Army GVSC
• Brandon Perelman, Gregory Gremillion, US Army Research Lab (ARL)

Industry

• Thomas Anthony, Analytical AI, LLC

Project began Q4 2022.

The overall motivation of this project is to determine how to render the autonomous assistant a trusted, reliable member of a warfighter crew, efficiently and effectively facilitating peak performance for military missions. To meet this goal, we will introduce the human-in-the-loop for shared autonomy and safety.

The overarching objective of this project is to develop an ecologically valid, high-fidelity experimental simulation to assess crew performance and communication under varying levels of AI assistance. The experimental procedure will determine (1) the impact of automated assistance on crew performance, situational awareness, and communication; (2) the minimal accuracy and precision of AI object detection and driving assistance necessary to fully support a ground vehicle crew; and (3) the optimal visual behavior of the commander, lookouts, and support drivers to mitigate risk and enhance driving performance.

The project will identify specific optimal performance and risk mitigation targets as well as benchmarks in automated assisted military missions. The chief contribution of the proposed project will be to evaluate the impact on mission efficacy when automated assistance is introduced to a 6-member crew. As part of the planning and setup portion of the proposed project, simulator and project specific AI algorithms for real time object detection will be developed to recognize significant objects presented by the simulated environment and alert mission personnel to their presence. As part of the data analysis portion of the proposed project, automation algorithms will be developed to expedite research processes (e.g., data reduction, image processing, qualitative analyses). Further, analytical prediction models will be developed to determine (1) the ideal number of automated assistance technologies used, and (2) the minimal real-world efficacy of automated assistance required to benefit mission objectives. Similarly, models quantifying the amount of cognitive resources freed by the automated assistance will be developed, based on the benefits of the assistance to (1) situational awareness, (2) driving performance, and (3) team communication.

#2.A101