Particle.news
Download on the App Store

JPL’s ERNEST Rover Validates Active‑Suspension Autonomy in Long‑Distance Desert Trials

March tests showed reinforcement‑learning autonomy and mobility gains that could shape higher‑speed, long‑range rovers for the Moon and Mars.

Overview

  • The prototype, which completed a seven‑day desert campaign in March 2026, drove about 16 miles (26 kilometers) over 37 hours with minimal human intervention.
  • ERNEST uses an active suspension, two powered front joints and four steerable wheels to lift individual wheels, change posture, and execute new gaits such as squirming and wheel‑walking.
  • JPL trained the rover’s decision making with reinforcement learning in a high‑fidelity simulator by running thousands of virtual driving hours to match hardware responses across terrain types.
  • The rover reached top speeds near 0.6 mph (1 km/h), roughly ten times faster than current Mars rovers, supporting concepts for longer “science road trips” and faster surface traversal.
  • Work on ERNEST began in 2022 as internal JPL R&D and is now funded by NASA program offices as engineers integrate active‑suspension control with longer‑range navigation and study scaling for lunar or Mars missions.