A first significant result of the use of Artificial Intelligence techniques for autonomously supervising a spacecraft has been obtained in May 1999 by NASA Ames and JPL with the Remote Agent project. An Artificial High Level Vision Agent for the Interpretation of the Operations of a Robotic Arm. Symposium on Artificial Intelligence, Robotics and Automation in Space, The Netherlands, 1999[Finzi et al.
Validating the ds1 remote agent experiment
Smith, William Taylor and Yu-wen Tung This paper describes the validation of the Remote Agent Experiment.
Gamble Jr., Bob Kanefsky, James Kurien, William Millar, Nicola Muscettola, Kanna Rajan, Nicolas Rouquette, Benjamin D.
The paper demonstrates that formal methods tools can find concurrency errors that indeed lead to loss of spacecraft functions, even for the complex software required for autonomy.
Second, it describes progress in automatic translation and abstraction that eventually will enable formal methods tools to be inserted directly into the aerospace software development cycle.
This demonstration included both nominal operations with goal-oriented commanding and closed-loop plan execution, and fault protection capabilities with failure diagnosis and recovery, on-board replanning following unrecoverable failures, and system-level fault protection.
A primary goal of this experiment was to provide an onboard demonstration of spacecraft autonomy.A second quick-response "cleanroom" verification effort found the concurrency error in a short amount of time.The error was isomorphic to one of the concurrency errors found during the first verification effort.On the other side, there is the need of simplifying the control of the mission by allowing the possibility that the scientists give a high level goal description to the robot and the (autonomous) robot is able to perform the assigned task without requiring low level instructions from the humans. In order to fulfill such ambitious goals, it is necessary to combine different techniques and methodologies, but in any case the adoption of Artificial Intelligence methodologies seems to be necessary [Doyle 97, Muscettola et al. In fact, tasks as planning, scheduling, diagnosis and reconfiguration al require reasoning capabilities and an explicit representation of the knowledge about the robot, the task and the environment. These goals were achieved by successfully integrating the Remote Agent with the Deep Space 1 #ight software, developing a layered testing approach, and taking various steps to gain the con#dence of the spacecraft team.