NASA Uses AI to Navigate Mars Rover Perseverance – A Historic First
The United States’ National Aeronautics and Space Administration (NASA) has achieved a landmark in Mars exploration. For the first time, NASA successfully employed artificial intelligence (AI) to chart a navigational course for the Perseverance rover on the Red Planet, marking a new chapter in the efficiency of space missions.
AI Claude Trial Runs Successfully
Two key demonstrations took place on December 8 and 10, 2025, where the AI model Claude, developed by Anthropic, was tasked with planning routes for Perseverance. The team at NASA’s Jet Propulsion Lab (JPL) double-checked the AI-generated paths to ensure safety and prevent the rover from encountering hazards.
The results were highly satisfactory. Perseverance covered nearly 1,500 feet (about 457 meters) during the two journeys without any recorded issues. This process mirrored the approach typically used by human operators, with Claude receiving satellite imagery and data from NASA’s Mars Reconnaissance Orbiter, just as JPL scientists do.
The paths generated by the AI were slightly modified by NASA before being sent to Perseverance, which then autonomously drove along the designated route. NASA Administrator Jared Isaacman expressed his appreciation for the achievement.
“This demonstration shows how far our capabilities have come and expands how we will explore other worlds,” Isaacman stated. “Technology like this can help missions operate more efficiently, respond to challenging terrain, and improve scientific outcomes as we travel farther from Earth. This represents a strong example of a team carefully and responsibly applying new technology in real operations.”
Efficiency and the Future of Exploration
While AI is often associated with lower-quality content, it has proven highly beneficial in the scientific field. Manually planning routes requires significant time to analyze years of imagery and data, and to plot safe waypoints.
Currently, waypoints are typically set no more than 330 feet (around 100 meters) apart, meaning Perseverance explores Mars step-by-step. For example, Perseverance’s journey out of Jezero Crater in 2024 took 3.5 months and climbed a total of 1,640 feet (approximately 500 meters) vertically. By December 2025, the rover had travelled a total of around 25 miles (approximately 40 km) in about four years.
Vandi Verma, a robotics expert at JPL, explained that the primary goal is to enable Perseverance and other Mars rovers to travel much greater distances while “minimizing operator workload.” Verma also added that AI could be used to identify interesting features on the planet, saving the time of human science teams by eliminating the need to manually examine “large volumes of rover imagery.”
A video of the December 10th journey can be viewed on NASA’s YouTube channel, condensed into a 52-second video. In the video, the route planned by Claude is shown in magenta, while the actual path taken by the rover is orange. NASA scientists only needed to make minor adjustments to the AI-planned route.
Frequently Asked Questions
What AI model was used to navigate Perseverance?
The AI model used was Claude, developed by Anthropic.
How far did Perseverance travel during the AI trial runs?
Perseverance travelled almost 1,500 feet (approximately 457 meters) during the two trial runs on December 8 and 10, 2025.
What is the potential benefit of using AI for rover navigation?
AI can help missions operate more efficiently, respond to challenging terrain, and improve scientific outcomes, while also minimizing the workload for human operators.
As NASA continues to refine this technology, could we see a future where rovers independently explore vast regions of Mars, accelerating the pace of scientific discovery?