Adaptive mission planning for coupled human-robot teams

Abstract

Existing approaches to multi-agent task allocation and scheduling do not extend well to missions in which humans coordinate closely with robotic teammates due to the dynamic and stochastic nature of human performance. This work develops fast algorithms that integrate humans into the planning problem to produce allocations for tightly coupled human-robot teams. Humans are treated as dynamic agents, and feedback from the realized performance of the team is leveraged to adapt human and environment models in real-time. Simulations of complex multi-agent missions demonstrate that the resulting plans outperform those of alternative strategies.