Making one's way through a crowded train station or airport is a serious challenge for people with visual impairments. A new study from Orebro University (Sweden) teaches robots to follow the "unwritten rules" of human movement and find a path through a dense crowd. In the long term, the technology could become a digital complement to guide dogs.

Today's navigation robots perceive every person as an obstacle to be avoided. When the crowd becomes too dense, they can run into the so-called "robot freezing problem": there is no clear path, and the machine stops completely.

Researchers at Orebro University have developed a new navigation system that allows robots to move through a busy public environment. The method is based on a different approach: people in a crowd do not stop. Instead, we adapt, keep to one side and move with the flow.

"If a robot is to guide a person with a visual impairment through a crowded terminal, it cannot stop completely every time someone walks past. It has to move naturally and follow the flow," says Yufei Zhu, who recently defended his doctoral thesis in computer science.

The work was published in the proceedings of the 2026 IEEE International Conference on Robotics and Automation (ICRA).

In his work, Zhu developed so-called "dynamics maps." Instead of mapping only stationary objects - walls and doorways - the robot learns how people usually move in a particular place at different times of day.

"You can think of it like this: a newcomer arrives in a city, observes how the locals behave and then copies them. By studying movement flows, the robot learns the patterns and can predict where people are heading," Zhu explains.

The technology, called HiCrowd, means the robot no longer perceives the crowd as a threat or an obstacle. Instead, it uses people's movement patterns as a guide.

The method was tested in real conditions at the Miraikan science museum in Tokyo and at Expo 2025 in Osaka. A robot the size of a small suitcase had to make its way through dense flows of visitors.

"Once the robot crossed a crowded hall behind a group of people. When the group turned right, the robot followed it, even though it could not yet see why. A few seconds later the reason became clear: a large crowd was coming the other way. By following those ahead, the robot safely passed the oncoming flow without stopping once," Zhu recounts.

The tests also showed that people around did not appear alarmed by the robot's appearance and did not try to keep away from it. The machine fitted into the flow so naturally that people around perceived it as an ordinary pedestrian.

One possible future application of the technology is helping people with visual impairments. A guide robot leading a person through a noisy train station or airport terminal must move smoothly, without abrupt movements and constant stops.

"At a train station there are clear habits and movement patterns. With our methods, a guide robot could lead a person with a visual impairment so that they move naturally with the flow rather than fighting it," Zhu says.

The research was carried out at Orebro University under the supervision of Professor Martin Magnusson in close collaboration with researchers from the Miraikan accessibility laboratory in Tokyo and the company Bosch in Germany. Thanks to this international partnership, the dynamic mapping technology was applied in a wide variety of conditions - from warehouse robots in an industrial environment to service robots moving among large groups of museum visitors.

Source: Tech Xplore