Russian Scientists Teach Robots to Navigate Like Humans

A new memory-inspired algorithm allows robots to move without GPS or detailed maps—marking a leap toward cognitive navigation.

Researchers from the Russian Academy of Sciences and the Moscow Institute of Physics and Technology have developed a novel robot navigation system based on how human memory works. Unlike conventional systems that rely heavily on GPS or detailed environmental maps, the new approach uses a cognitive-style algorithm.

The method allows robots to identify key landmarks and build connections between them—enabling them to navigate with fewer computing resources and plan routes more efficiently. It mimics the human brain’s way of processing spatial information.

Initial tests in virtual 3D environments and real-world trials showed the system's robustness. Despite sensor noise, robots were able to build accurate, coherent map-schemes.

The system, dubbed PRISM-TopoMap, updates its internal map dynamically as the robot moves. Researchers plan to expand its capabilities further, including teaching robots to recognize room types—a step toward more intelligent indoor navigation.