Classifying Autonomy for Mobile Space Exploration Robots
Sylvain Joyeux, Jakob Schwendner
In Proceedings of the 10th International Symposium on Artificial Intelligence, Robotics and Automation in Space, (iSAIRAS-10), 29.8.-01.9.2010, Sapporo, o.A., Sep/2010.
Robotic exploration missions to celestial bodies in our solar system provide us with the necessary knowledge to prepare for human presence on the moon or other planets, and could have the ability to give vital information about our own origins. Direct control over the mobile robot systems is only possible in particular cases, and the ability of the systems to take decisions autonomously increased in importance with more complex mission scenarios and targets further away. The ability to classify and benchmark autonomy is important for the design, evaluation and comparison of such systems. One of the problems is that existing schemes to classify autonomy are either too generic, or don't fit the requirements for mobile space exploration systems. The existing schemes for space engineering and other domains are discussed. Further, existing and planned robotic exploration missions are analysed towards their autonomy content. A new scheme for the classification of autonomy in the given domain is drafted and the existing systems classified. The evaluation of performance levels for autonomy is discussed, and initial ideas on how this could be feasible given. The paper is considered a starting point for further work in the area of classification and benchmarking of autonomy for mobile space exploration robots.