Unmanned vehicles learned to enter difficult turns at speed and brake on the ice

Existing AI systems for controlling self-driving cars are doing a good job in

under normal circumstances, but in the event of emergency situations, such as black ice, the need to perform a maneuverAt high speeds and others, mistakes often occur.

The researchers have created a system that will allowUAVs confidently deal with emergency situations. The testing of the algorithm installed in the Volkswagen GTI and Audi TTS cars took place on an ice-covered test track near the Arctic Circle and on the Thunderhill racetrack in California.

As a result, the cars managed to successfully overcome sharp turns on the ice-covered road, as well as fit into the turn at high speed and slow down.

Our project is designed to improve the safety of autonomous vehicles – we want them to successfully cope with their tasks in a variety of scenarios, fromNormal driving on asphalt with a high coefficient of friction to fast, low-friction driving on ice and snow.

Nathan Spielberg, lead author of the study

Earlier it was reported that unmanned trams will appear in Russian millionaire cities, where residents will be able to get quick medical care.