Nicola Muscettola, Ph.D.
NASA Ames Research Center
Wednesday, December 14, 2005
Pacific Forum – 12:00 p.m.
The Mars Exploration Rover (MER) shows that it is possible to produce scientific breakthroughs using robotic systems capable of a moderate amount of autonomy. Challenges to autonomous operations will increase substantially as we tackle the exploration of more dynamic environments, such as Titan's atmosphere or Europa's Ocean. The core challenges are: dealing with sporadic communication with mission control, ensuring safe operation in the presence of anomalous conditions, reacting swiftly to serendipitous science opportunities, restructuring subsystem interactions to deal with performance degradation or failure, and using limited resources effectively.
This seminar illuminates the use of Intelligent Distributed Execution Architecture software (IDEA) for the purpose of space exploration. This framework tackles challenges by making use of on-board and off-board planning as the primary computational engine for supervisory control. IDEA has been deployed in several robotics demonstrations including a space interferometer prototype and a multi-Unmanned Aerial Vehicle system (UAV) for the Department of Defense Advanced Research Projects Agency (DARPA).
This presentation will describe two rover applications. The first application is the Collaborative Decision Systems (CDS) human-robotics system which demonstrated the use of two rovers that operated from a predetermined plan and responded to serendipitous commands issued by a suited astronaut. The “Life in the Atacama” project, conducted in northern Chile's Atacama Desert is an example of the second rover application. During this project scientists have utilized Carnegie Mellon University’s multi-UAV system, the Zoe rover, to study life in a hostile environment. The results from this project may ultimately enable multi-UAV systems to look for life on Mars.