Using supervised learning to compensate for high latency in planetary exploration

Planetary exploration utilizes orbital, aerial, surface and potentially sub-surface vehicles at remote locations. At present, many of these vehicles must be teleoperated or commanded from Earth, requiring data to be transmitted over significant distances (taking several minutes or longer). This problem gets progressively more pronounced as vehicles are operated further and further from the Earth. While missions to send humans to deep space and increased autonomy both prospectively present partial solutions to this challenge, human controllers may wish to operate the vehicle with an experience more akin to in-situ exploration. Typically, using remotely operated vehicles for exploring planets entails dealing with a high amount of latency. This paper explores the use of supervised learning for rovers that would reduce the amount of data that needs to be sent to control the vehicle and provide data to satisfy future user needs.