A Probabilistic Approach to Robust Execution of Temporal Plans with Uncertainty

Abstract

In Temporal Planning a typical assumption is that the agent controls the execution time of all events such as starting and ending actions. In real domains however, this assumption is commonly violated and certain events are beyond the direct control of the plan’s executive. Previous work on reasoning with uncontrollable events (Simple Temporal Problem with Uncertainty) assumes that we can bound the occurrence of each uncontrollable within a time interval. In principle however, there is no such bounding interval since there is always a non-zero probability the event will occur outside the bounds. Here we develop a new more general formalism called the Probabilistic Simple Temporal Problem (PSTP) following a probabilistic approach. We present a method for scheduling a PSTP maximizing the probability of correct execution. Subsequently, we use this method to solve the problem of finding an optimal execution strategy, i.e. a dynamic schedule where scheduling decisions can be made on-line.