A computer program for use aboard a scientific- exploration spacecraft autonomously selects among goals specified in high-level requests and generates corresponding sequences of low-level commands, understandable by spacecraft systems. (As used here, "goals" signifies specific scientific observations.) From a dynamic, onboard set of goals that could oversubscribe spacecraft resources, the program selects a non-oversubscribing subset that maximizes a quality metric. In an early version of the program, the requested goals are assumed to have fixed starting times and durations. Goals can conflict by exceeding a limit on either the number of separate goals or the number of overlapping goals making demands on the same resource.
The quality metric used in this version is chosen to ensure that a goal will never be replaced by another having lower priority. At any time, goals can be added or removed, or their priorities can be changed, and the "best" goal will be selected. Once a goal has been selected, the program implements a robust, flexible approach to generation of low-level commands: Rather than generate rigid sequences with fixed starting times, the program specifies flexible sequences that can be altered to accommodate run time variations.
This program was written by Gregg Rabideau, Steve Chien, and Ning Liu of Caltech for NASA's Jet Propulsion Laboratory.
This software is available for commercial licensing. Please contact Karina Edmonds of the California Institute of Technology at (626) 395-2322. Refer to NPO-44541.
This Brief includes a Technical Support Package (TSP).
Providing Goal-Based Autonomy for Commanding a Spacecraft
(reference NPO-44541) is currently available for download from the TSP library.
Don't have an account?
Overview
The document titled "Providing Goal-Based Autonomy for Commanding a Spacecraft" from NASA's Jet Propulsion Laboratory discusses the development of high-level, goal-based autonomy for spacecraft operations. It addresses the challenges of managing conflicting goals and optimizing resource allocation on board spacecraft, particularly in scenarios where multiple objectives must be prioritized and executed dynamically.
The core problem tackled in the document is the need for a dynamic, on-board goal set that can adapt to changing mission requirements. This involves automated goal triggering, both on-board and from ground-based systems, allowing for flexible execution of command sequences. The document emphasizes the importance of a robust goal selection process, where goals can conflict based on priority and resource limitations. A prototype system, utilizing a Virtual Machine Language (VML) sequencer and a Goal Dispatcher (GD), is introduced to manage these goals effectively.
The GD maintains a list of requested goals, their priorities, and the constraints between them. As goals approach their execution time, the GD evaluates whether they remain the best option for execution, ensuring that the highest priority goals are addressed first. This system allows for the addition, removal, or modification of goals at any time, enhancing the spacecraft's ability to respond to real-time conditions and anomalies.
The document also highlights the significance of data volume in scientific requests, noting that available data can influence the duration and complexity of tasks. For instance, if a high-priority data downlink is required due to an anomaly, the system can dynamically adjust its goals to accommodate this need, replacing lower-priority science goals as necessary.
Additionally, the concept of adjustable autonomy is discussed, which allows for varying levels of autonomy based on mission requirements and the evolving confidence in the spacecraft's systems. This flexibility is crucial for long-term missions, where risk tolerance may change over time.
Overall, the document outlines a comprehensive approach to goal-based autonomy in spacecraft commanding, emphasizing the need for dynamic goal management, flexible execution, and the ability to adapt to changing mission conditions. This innovative framework aims to enhance the efficiency and effectiveness of space missions, ultimately contributing to the success of NASA's exploration objectives.
