Map Perseverance routes using AI

Map Perseverance routes using AI

explanation

Annotated images from NASA’s HiRISE (High-Resolution Imaging Science Experiment) camera aboard NASA’s Mars Reconnaissance Orbiter Image show the AI-planned route and the actual route that NASA’s Mars rover Perseverance followed while traveling 807 feet (246 meters) on December 10, 2025, the 1,709th Martian day (sol) of its mission. The run was the second of two demonstrations, the first of which took place on December 8, to show that generative artificial intelligence can be incorporated into the rover’s route planning.

The magenta line shows the path the rover’s wheels will take when following the AI-processed waypoints shown by the magenta circles. (A waypoint is a fixed location where the rover receives a new set of instructions.) The orange line is based on data downlinked after the drive is completed and shows the actual path taken by the rover. At the start of the route, the short, bold segment of the blue line in the upper right corner indicates the portion of the drive determined by the mission’s rover driver and based on images taken by the rover of the surface ahead. The surface area inside the light green box is called the “keep-in zone.” Perseverance’s self-driving software can only select routes within these zones.

This graphic was generated using Hyperdrive, part of the software suite used to plan the rover’s drives and manage the large influx of engineering data from the Perseverance rover.

NASA’s Jet Propulsion Laboratory is managed by the California Institute of Technology and manages the construction and operations of the Perseverance spacecraft.

For more information about Perseverance, visit science.nasa.gov/mission/mars-2020-perseverance/.

The University of Arizona in Tucson operates HiRISE, built by BAE Systems in Boulder, Colorado. JPL manages SMD’s Mars Reconnaissance Orbiter.