The first complex for monitoring the state of roads using neural networks has been launched in Russia

March 2, 2021 it was announced the launch in Tula of a mobile complex for monitoring the state of roads using neural networks. This is the first such decision in Russia, says its developer, Rostelecom.

The software and hardware complex (PAK) installed on the customer's car consists of video surveillance cameras and an artificial neural network module. The system is able to automatically record deviations in the state of road infrastructure and adjacent territory: a highway, curbs, road signs, traffic lights, lighting supports and others. With the help of a mobile complex, you can determine signs of pollution, the presence of paint, stickers and inscriptions on the surface of information signs, control the overflow of ballot boxes near public transport stops and other violations.

The first complex for monitoring the state of roads using neural networks from Rostelecom has been launched in the Russian Federation

Equipped with video cameras, the transport records malfunctions. Data on the state of infrastructure and facilities are recorded in PAK memory, as well as in real time are transferred to the customer's personal account located in the Rostelecom cloud. The information contains the date, time and place of fixation, as well as a photograph of the object indicating the type of deviation.

The director of the branch of PJSC Rostelecom in the Tula and Ryazan regions, Ivan Anashkin, called the hallmark of the new complex that continuous video shooting of objects takes place from the vehicle, and the neural network on-line automatically analyzes their condition without recording and storing video.

According to Anashkin, by the beginning of March 2021, such equipment was installed only on one car, but in the future it is planned to place it on urban public transport and communal equipment to survey yard areas.^[1]

Notes