Intention Communication and Hypothesis Likelihood in Game-Theoretic Motion Planning

Game-theoretic motion planners are a potent solution for controlling systems of multiple highly interactive robots. Most existing game-theoretic unrealistically assume a priori objective function knowledge is available to all agents. To address this, we propose fault-tolerant receding horizon planner that leverages inter-agent communication with intention hypothesis likelihood. Specifically, robots communicate their which incorporates intentions. A discrete Bayesian filter designed infer the objectives in real-time based on discrepancy between observed trajectories and predicted solutions non-cooperative games under hypotheses. In simulation, consider three safety-critical autonomous driving scenarios overtaking, lane-merging intersection crossing, demonstrate our planner's ability capitalize alternative hypotheses generate safe presence faulty transmissions network.