Probabilistic temporal planning with uncertain durations

Few temporal planners handle both concurrency and uncertain durations, but these features commonly co-occur in real-world domains. In this paper, we discuss the challenges caused by concurrent, durative actions whose durations are uncertain. We present five implemented algorithms, including ΔDURprun, a planner guaranteed to find the optimal policy. An empirical comparison reveals that ΔDURexp, our fastest planner, obtains orders of magnitude speed-up compared to ΔDURprun--with little loss in solution quality. Importantly, our algorithms can handle probabilistic effects in addition to stochastic durations, and they are effective even when duration distributions are multi-modal.