APSAdaptive Problem Solving for Large Scale Scheduling and Resource Allocation Problems task involves the development of Machine Learning Methods to automatically find effective domain-specific scheduler control strategies. It is applied to the domains of spacecraft commanding and to Deep Space Network scheduling problems.
CLEaRClosed Loop Execution and Recovery is an integrated planning and execution framework for autonomous control of robotic entities. The CLEaR system currently utilizes CASPER and TDL, and focuses on on the use of both near-term reactive behavior and long-term deliberative decision making.
Benchmark Problem DomainsThe benchmark problem domains were developed and released in response to the disconnect between the research and application communities in planning and scheduling. The two actual space domains, DATA-CHASER and CX-1, developed at Colorado Space Grant, contain multiple levels of difficulty and simulators corresponding to these levels, in the hopes the the planning and scheduling community will direct research towards more real-world problems.
3CSThree Corner Satellite is a demonstration of stereo imaging, formation flying and innovative command and data handling, including on-board autonomy + Read More
CX1Citizen Explorer is a small earth orbiting satellite built and managed by the Colorado Space Grant Consortium.
DS-1Deep Space 1 The first deep space flight of the New Millenium program will feature advanced software for autonomous operation.
MAMMModified Antarctic Mapping Missison The ASPEN planning system automated the mission planning process and provided a fast replanning capability for responding to anomolies during operations.
Autonomous Rover Command GenerationA proof-of-concept prototype for automatic generation of validated rover command sequences from high-level science and engineering activities.
Rocky-7 Rover Science PlanningAn intelligent science tool for planetary rover operations.
Deep Space Network Operations
DSSCDeep Space Station Controller is an extension to the work performed for the Deep Space Terminal (DS-T) task, part of the Deep Space Network, in the area of track automation. This work utilizes the CLEaR system to provide the capability for robust dynamic desision making and execution management for autonomous DSN ground station operations.
Autonomous Aerial Vehicles
UAVsUnmanned Air Vehicles consist of integrating JPL planning, diagnostics, and prognostics systems as part of the control architecture for UAVs.
ASIPAutomated SAR Image Processing system uses Artificial Intelligence Planning techniques to automate most fo the steps in image processing of synthetic aperature radar (SAR) images to satisfy science requests
MVPMultimission VICAR Planner uses Artificial Intelligence Planning techniques to automatically synthesize executable image processing processing procedures to satisfy science requests.
Technical QuestionsSteve Chien
4800 Oak Grove Dr.
Pasadena, CA 91109-8099
steve.chien at jpl.nasa.gov
+ Home Page