Navigation in Partially Observed Dynamic Roadmaps

We consider robot navigation in environments consisting of a known static map, but where dynamic obstacles of vary- ing and unknown lifespans appear and disappear over time. We describe a roadmap-based formulation of the problem that takes the sensing and transition uncertainty into account, and an efficient online planner for this problem. The planner naturally displays behaviors such as persistence and obstacle timeouts, and is able to make inferences about obstacle types even with impoverished sensors.