Knowledge Acquisition for the Onboard Planner of an Autonomous Spacecraft

Deep Space One (DS1) will be the first spacecraft to be controlled by an autonomous closed loop system potentially capable of carrying out a complete mission with minimal commanding from Earth. A major component of the autonomous flight software is an onboard planner/scheduler. Based on generative planning and temporal reasoning technologies, the planner/scheduler transforms abstract goals into detailed tasks to be executed within resource and time limits. This paper discusses the knowledge acquisition issues involved in transitioning this novel technology into spacecraft flight software, developing the planner in the context of a large software project and completing the work under a compressed development schedule. Our experience shows that the planning framework used is adequate to address the challenges of DS1 and future autonomous spacecraft systems, and it points to a series of open technological challenges in developing methodologies and tools for knowledge acquisition and validation.