CL 01-0573

Background

An autonomous spacecraft must balance long-term and short-term considerations. It must perform purposeful activities that ensure long-term science and engineering goals are achieved and ensure that it maintains positive resource margins. This requires planning in advance to avoid a series of shortsighted decisions that can lead to failure. However, it must also respond in a timely fashion to a somewhat dynamic and unpredictable environment. Thus, spacecraft plans must often be modified due to fortuitous events such as early completion of observations and setbacks such as failure to acquire a guidestar for a science observation.

CASPER (Continuous Activity Scheduling Planning Execution and Replanning) uses iterative repair to support continuous modification and updating of a current working plan in light of changing operating context.

AI Technology

• Planning and Scheduling
• Planning and Execution
• Replanning

Problem

Traditional batch oriented models of planning have shortcomings for spacecraft control. Constructing a plan from scratch can be a computationally intensive process and onboard computational resources are typically quite limited, so that it still may require considerable time to generate a new operations plan. As a data point, the planner for the Remote Agent Experiment (RAX) flying on-board the New Millennium Deep Space One mission takes approximately 4 hours to produce a 3 day operations plan. RAX is running on a 25 MHz RAD 6000 flight processor and uses roughly 25% of the CPU processing power. While this is a significant improvement over waiting for ground intervention, making the planning process even more responsive (e.g., on a time scale of seconds) to changes in the operations context, would increase the overall time for which the spacecraft has a consistent plan. As long as a consistent plan exists, the spacecraft can keep busy working on the requested goals.

Impact

Making the planner more timely in its responses has a number of benefits:

• The planner can be more responsive to unexpected (i.e., unmodelable) changes in the environment that would manifest themselves as updates on the execution status of activities as well as monitored state and resource values.
• The planner can reduce reliance on predictive models (e.g., inevitable modeling errors), since it will be updating its plans continually.
• Fault protection and execution layers need to worry about controlling the spacecraft over a shorter time horizon (as the planner will replan within a shorter time span).

Status

CASPER is being used in a range of projects including autonomous spacecraft, autonomous rovers, ground communications station automation, and uninhabited aerial vehicles.

Description

To achieve a higher level of responsiveness in a dynamic planning situation, we utilize a continuous planning approach and have implemented a system called CASPER (for Continuous Activity Scheduling Planning Execution and Replanning). Rather than considering planning a batch process in which a planner is presented with goals and an initial state, the planner has a current goal set, a plan, a current state, and a model of the expected future state. At any time an incremental update to the goals or current state may update the current state of the plan and thereby invoke the planner process. This update may be an unexpected event or simply time progressing forward. The planner is then responsible for maintaining a consistent, satisficing plan with the most current information. This current plan and projection is the planner's estimation as to what it expects to happen in the world if things go as expected. However, since things rarely go exactly as expected, the planner stands ready to continually modify the plan. Current iterative repair planning techniques enable incremental changes to the goals and the initial state or plan and then iteratively resolve any conflicts in the plan. After each update, its effects will be propagated through the current projections, conflicts identified, and the plan updated (e.g., plan repair algorithms invoked).

Applications

• Distributed S/C
• Citizen Explorer
• CLEaR
• DS-T
• Planetary Rover Operations
• Distributed Rovers
• New Millennium Earth Orbiting 1 (EO1)
• Highly Reusable Space Transportation
• Modified Antarctic Mapping Mission (MAMM)

Publications

Contacts

JPL Technical Contact:	Dr. Steve Chien Steve.Chien at jpl.nasa.gov 818.393.5320
Software Licensing:	http://download.jpl.nasa.gov

Sponsors

NASA Code SM
Autonomy program, Dave Atkinson (JPL) managing.
Telecommunications and Mission Operations Technology (TMOT)
Peter Shames, Managing

Also Sponsored By:
Directors Research Discretionary Fund
Jet Propulsion Laboratory
California Institute of Technology