AI Is Casting a Net for Scooter Rule-Breakers
From 2020 to 2025, the total number of traffic crashes involving personal mobility devices (PMDs), including electric scooters, in Russia increased eightfold, from 410 to 3,275. Electric scooter riders are at fault in 68% to 71% of these crashes, largely because they violate traffic rules. Artificial intelligence is now being deployed to monitor compliance.
In Yoshkar-Ola, NtechLab, a technology partner of Rostec, has launched an AI pilot project to identify electric scooter riders who violate bans on riding in pedestrian zones.
An AI That Keeps a Close Watch
The project does not require new infrastructure. Cameras are already in place to monitor roads, parks and public squares. The system taps into existing video feeds and begins analyzing them, with electric scooters now among the targets of the city’s smart cameras.
Yoshkar-Ola has about 20 zones where riding electric scooters is prohibited, including the Brugge Embankment, Voskresenskaya Embankment, Tsargradskaya Square, Italian Park and other public spaces. These restricted areas are mapped digitally. When the system detects a scooter rider inside one of them, it flags a potential violation and forwards the information to the responsible city services. Shared scooters typically display a number or have distinctive design on their body. The system can read the number or identify the operator from the scooter’s appearance. That makes it possible to determine which company owns the scooter and notify both city authorities and the operator, which can then take action against the rider.
Testing the Technology Across Russia
In recent years, Russia has accumulated a growing portfolio of video analytics projects in which neural networks analyze camera footage and automatically detect events on roads and in urban areas. MTS launched a camera-based AI monitoring system for electric scooters in Sirius. It detects riders crossing pedestrian crosswalks without dismounting, riding in prohibited areas, carrying two people on one scooter and parking improperly. The system sends the information to the scooter-sharing operator, which then notifies the user. During the first six months of operation, the system issued more than 600 warnings. According to the company, more than 90% of those warned tried not to repeat the violations.
“Even after just six months of testing, the measures we introduced have proved highly effective. Improper scooter riding is a problem in many cities, so this technology could be in strong demand in other Russian regions as well,” said Dmitry Plishkin, head of the Sirius administration.
Urbantech’s Azimut systems have demonstrated high accuracy in automatically detecting scooter-related violations. Netvision has developed a platform that analyzes footage from city cameras directly on the device, without transmitting the entire video stream to servers. It also gives city administrations a dashboard featuring traffic heat maps and hotspots where violations occur. According to the company, 88% of electric scooter violations involve riding across pedestrian crosswalks. That finding emerged after the technology was deployed in the Samara region.
Scooter operator Whoosh takes a fundamentally different approach to catching rule-breakers. Instead of analyzing camera footage, its system uses data from the scooter’s own sensors. It analyzes more than 50 ride parameters to detect when two people are riding on a single scooter. The system immediately alerts the user and automatically reduces the scooter’s speed to a level the company describes as “uncomfortable and inefficient.” According to company statistics, users then either end the ride or switch to two separate scooters. In a single day, the system detected more than 400 such rides in St. Petersburg and more than 800 in Moscow. The technology operates in every city where Whoosh is present in Russia, Belarus and Kazakhstan.
A Powerful Tool for Micromobility
In Moscow, nearly 400 cameras detect scooter-related violations. At the same time, the number of crashes involving electric scooters has fallen by half, even as the city’s shared-scooter fleet has grown to 60,000 vehicles. Experience from regions that have introduced similar technologies suggests that automated monitoring can make pedestrian areas safer, reduce conflicts between pedestrians and PMD users, and help city services respond to violations more quickly.
Yoshkar-Ola is considering automatically issuing fines for riding scooters in prohibited areas and improper parking. Before such a system can be introduced, however, authorities will need to establish a legal basis for it and define procedures for recording violations and confirming the identity of the person responsible.
A major advantage of these systems is that they can use existing surveillance cameras rather than requiring cities to build separate infrastructure.
