Skip to main content
arc logo
Automotive Research Center
hero image
Back to all projects

Human-Autonomy Interaction

Annual Plan

Optimal Distribution of Tasks in Human-Autonomy Teams

Project Team

Principal Investigator

Bogdan Epureanu, University of Michigan

Government

Victor Paul, Jillyn Alban, U.S. Army GVSC

Industry

Mert Egilmez, Veoneer-Nissin

Student

Amin Ghadami (Post-Doc), University of Michigan

Project Summary

Project started September 2019.

Recent developments in autonomy offer several opportunities that might lead to a paradigm shift in several domains including military ground systems. In this new paradigm, humans and autonomous assets (such as vehicles) can collaborate during operation as a team to improve capabilities of existing human operated systems. In a team, autonomous assets as well as humans have inherent limitations in distinct mission-related attributes. For example, autonomous assets are limited in terms of real-time computational power, amount and accuracy of sensor data and accuracy of decision-making capabilities of artificial intelligence methods. At the same time, humans are limited in terms of cognitive loads, fatigue and reaction time.

A fundamental research challenge emerging is how to make complex decisions and perform the corresponding set of actions synergistically when humans and autonomous systems work together to achieve superior capability than the sum of these two taken alone. This requires an efficient distribution of the tasks each decision-maker can perform better relative to the others.

The effort in Year 1 will focus on creating simplified models of human and autonomous asset decision-making and using these models to coordinate a set of tasks for a complex mission with given team size and composition. These models will be simple enough to explore various task distributions and coordination strategies but will also include the right attributes to capture the dynamics of human-autonomous asset collaboration. The effort in Year 2 will create high-fidelity AI processes with state-of-the-art methods to replace the simplified model of autonomous assets and optimize the task distribution between humans and autonomous members. The high-fidelity AI processes will allow exploring the limitations and the benefits of existing AI methods and their impact on the team decision-making. The effort in Year 3 will replace the model of human decision making with real people in the loop in virtual environments (such as video games). These environments will allow exploring the human factors and the creative power of human cognition. The results of this study can change the way we conceive the human-autonomous asset collaboration today and inform the hardware/software design of more synergistic autonomous systems for the future.

Publications from Prior Work closely related to the proposed project:

  • Bayrak A. E., Collopy A., Papalambros P., Epureanu B., “Multiobjective Optimization of Modular Design Concepts for a Collection of Interacting Systems.” Structural and Multidisciplinary Optimization, 57(1), 83-94, 2017.
  • Bayrak A. E., Egilmez M. M., Kuang H., Li X., Park J. M., Hu J., Papalambros P. Y., Epureanu B.I., Umpfenbach E., Anderson E., and Gorsich D., “A System of Systems Approach to Strategic Feasibility of Modular Vehicle Fleets.” IEEE Transactions on Systems, Man, and Cybernetics: Systems, (99), 1-13, 2018.
  • Li X., Epureanu B. I., “Intelligent Agent-based Dynamic Scheduling for Military Modular Vehicle Fleets.” Proceedings of the 2017 Industrial and Systems Engineering Conference, Pittsburgh, PA, May 20-23, 2017.
  • Li X., Epureanu B. I., “Robustness and Adaptability Analysis of Future Military Modular Fleet Operation System.” Proceedings of the 2017 Dynamic Systems and Control Conference, Tysons Corner, VA, Oct 11-13, 2017.
  • Li X., Epureanu B. I., “An Agent-based Approach for Optimizing Modular Vehicle Fleet Operation.” International Journal of Production Economics, 2018 (to be submitted).
  • Li X., Epureanu B. I., “Analysis of Fleet Modularity in an Artificial Intelligence Based Defender-Attacker Game.” Journal of Operations Management, 2018 (to be submitted).