An autopilot for electric scooters has been created in Russia. The device was developed by students from the Ural Federal University and schoolchildren from the Sverdlovsk region. This was reported by the press service of the National Technology Initiative (NTI) Platform.
The system with built-in AI recognizes objects and allows determining the distance to them. Upon detecting obstacles in the way, the device automatically reduces the speed of the electric scooter to 5 km/h.
The autopilot is a system consisting of a computing module and two wide-angle cameras mounted on the scooter's handlebars. The module receives images from the cameras in real time, recognizes objects, and processes them. The objects can be pedestrians, road markings, signs and traffic lights, as well as possible obstacles for the scooter – curbs and walls. Then the module calculates the distance to these objects and, in case of a potential threat to the scooter rider, limits his possible actions.
The development has no analogues both in Russia and abroad. The authors of the project are counting on further development with entry into the market.
Earlier it was reported that the kicksharing service Whoosh is implementing computer vision in electric scooters.
Read materials on the topic:
Whoosh Implemented a Quick Start Function for Electric Scooters
Kicksharing Whoosh Localized the Production of Components for Electric Scooters in Russia
