Plan Authoring Based on Sketches, Advice and Templates

摘要

Artificial intelligence (AI) planning technology provides powerful tools for solving problems that require the coordination of actions in the pursuit of specified goals. To date, however, there has been limited success in transitioning this technology to significant applications in the commercial, military, and space sectors. A major obstacle to technology transfer lies with the lack of control available to potential users of planning systems. Al planning systems have traditionally been designed to operate as black boxes that take a description of a domain and a set of goals and automatically synthesize a plan for achieving the goals. Many potential consumers of planning technology require more user-centric tools that are designed to augment human skills rather than replace them. Both the military and space communities are showing tremendous interest in user-centric planning technology that combines plan authoring and automated decision aids. The main focus of this project was the development of the PASSAT system (Plan Authoring based on Sketches, Advice, and Templates), a mixed-initiative framework for developing complex, hierarchical plans. With its combination of interactive and automated capabilities, PASSAT enables a user to quickly develop plans that draw upon past experience encoded in templates but that are customized to individual preferences and the demands of the current situation.