Planning with Continuous Resources in Stochastic Domains

We consider the problem of optimal planning in stochastic domains with metric resource constraints. Our goal is to generate a policy whose expected sum of rewards is maximized for a given initial state. We consider a general formulation motivated by our application domain--planetary exploration--in which the choice of an action at each step may depend on the current resource levels. We adapt the forward search algorithm AO* to handle our continuous state space efficiently.