NASA's Perseverance Rover Completes First AI-Planned Drive on Mars
NASA's Perseverance rover has successfully executed its first Mars drives planned entirely by artificial intelligence, marking a significant advancement in autonomous space exploration.
NASA's Perseverance Mars rover has achieved a milestone by completing its first drives on another planet that were entirely planned by artificial intelligence. The tests, conducted on December 8 and 10, 2025, utilized generative AI to autonomously chart safe routes across the Martian terrain, a task traditionally handled by human mission planners.
"This demonstration shows how far our capabilities have advanced and broadens how we will explore other worlds," said NASA Administrator Jared Isaacman. He emphasized that autonomous technologies like this can enhance mission efficiency, adapt to challenging terrains, and increase scientific returns as missions venture farther from Earth.
The AI-driven navigation was developed by NASA's Jet Propulsion Laboratory (JPL) in collaboration with Anthropic, utilizing the company's Claude AI models. The AI analyzed high-resolution images from the HiRISE camera aboard NASA's Mars Reconnaissance Orbiter and terrain-slope data from digital elevation models to identify hazards such as boulder fields and sand ripples. It then generated continuous paths with waypoints for the rover to follow.
Before execution, the AI-generated plans underwent rigorous testing using JPL's digital twin of Perseverance, verifying over 500,000 telemetry variables to ensure compatibility with the rover's flight software. On December 8, Perseverance traveled 689 feet (210 meters), followed by an 807-foot (246-meter) drive on December 10, both guided by AI-planned routes.
Vandi Verma, a space roboticist at JPL and member of the Perseverance engineering team, noted that the fundamental elements of generative AI show promise in streamlining autonomous navigation for off-planet driving. She highlighted the potential for AI to handle kilometer-scale drives while minimizing operator workload and to identify interesting surface features by analyzing vast volumes of rover images.
This advancement addresses the challenge posed by the average 140 million-mile (225 million kilometers) distance between Mars and Earth, which results in significant communication delays. By enabling the rover to autonomously plan and execute routes, AI integration reduces reliance on Earth-based planning, paving the way for more efficient and responsive exploration of other planets.