Current path planning systems for NASA’s Mars rovers require extensive human input that can lead to wasted mission operations, due to communication delays and time-consuming manual analysis of data gathered from orbit and the rover.
The Intelligent Path Planner (IPP) allows for heterogeneous data sets to be aggregated in real time, enabling rovers to learn as they drive. The IPP uses the information it has gathered to influence the path it plans, while also adapting the path based on the terrain encountered. Machine learning teaches the IPP to associate different terrains with performance levels. IPP is targeted as a software upgrade to improve the efficiency, safety and scientific return of rover missions without the need for specialized sensors