High-Level Autonomy and Image prioritisation for long distance Mars Rovers

This paper describes an approach to high-level rover autonomy using on-board planning, image classification and prioritisation which has been carried out as part of a project called PRoViScout. This project seeks to prototype techniques which would allow remote rovers to autonomously acquire science data and additional sensor data based on pre-specified science goals. Such a capability will be required to support long range rover operations which are foreseen for future exploration missions. This paper focuses on the overall system architecture, smart image classification and on-board planning concepts.