Learning and planning for Mars Rover science
Tara A. Estlin, Rebecca Castaño, Robert Charles Anderson, Daniel M. Gaines, Forest Fisher, Michele Judd · 2003
Abstract unknown environments where unexpected conditions can With each new rover mission to Mars, rovers are traveling significantly longer distances. In some cases, distances are increasing by orders of magnitude from previous missions. This increase enables not only the collection of more science data, but causes a large rise in the number of new and different science collection opportunities. In this paper, we describe the OASIS system, which provides autonomous capabilities for dynamically pursuing these science-collection opportunities during long-range rover traverses. OASIS utilizes techniques from both machine learning and planning and scheduling to address this goal. Machine learning techniques are applied to analyze data as it is collected and quickly determine new science tasks and priorities on these tasks. Planning and scheduling techniques are used to alter the rover’s behavior so new science measurements can be performed while still obeying resource and other mission constraints. In addition to describing our system, we also discuss how we are testing OASIS, including the use of Mars rover prototypes and validation using data gathered from expert planetary geologists. 1