Human-Robot Team Task Scheduling for Planetary Surface Missions

Future manned space exploration missions will require collaborative activities with human astronaut and robotic agents. Proposed mission planning and execution architectures rely on reactive task planning, focusing on real-time information sharing between ground personnel, astronauts, and rovers. We present a planning/scheduling strategy that autonomously optimizes initial task plans/schedules over the human-robot team based on specified mission goals, but that also scales computationally to enable reactive replanning based on modifications to mission goals, revised resource utilization profiles, anomalies/emergencies, and astronaut directives. Collaborative exploration task models and search control algorithms are presented and evaluated in terms of computational complexity versus solution optimality. The implemented software was evaluated over teams of simulated and real rovers collaborating with a human companion. These tests demonstrate that the planning tool can reliably trade execution time for solution optimality when required and that the planner/scheduler is able to capably respond to detected anomalies.