Enabling fast flexible planning through incremental temporal reasoning

Abstract

In order for a team of autonomous agents to successfully complete its mission, the agents must be able to quickly re-plan on the fly as unforeseen events arise in the environment. This requires temporally flexible plans that allow the agent to adapt to execution uncertainties by not overcommitting on time constraints, and a continuous planner that replans at any point when the current plan fails. To achieve both of these requirements, planners must have the ability to reason quickly about timing constraints. This thesis provides a fast incremental algorithm, ITC, for determining the temporal consistency of temporally flexible plans. Additionally, the temporal reasoning capability of ITC is able to return the conflict or the nature of the inconsistency to the planner, such that the planner can resolve inconsistencies quickly and intelligently. The ITC algorithm combines the speed of shortest-path algorithms known to network optimization with the spirit of incremental algorithms such as Incremental A* and those used within truth maintenance systems (TMS). The algorithm has been implemented and integrated into a temporal planner, called Kirk. It has demonstrated an order of magnitude speed increase on cooperative air vehicle scenarios.