Recently, Xinbei dstrict of Changzhou launched AI-powered patrol vehicles to improve the efficiency of urban management.

The vehicles are equipped with 360-degree panoramic cameras and an AI-powered computer vision system. As they patrol the streets, images are transmitted in real time to the city appearance monitoring platform of the Xinbei District Bureau of Urban Management and Law Enforcement, where AI automatically identifies potential urban management issues and generates alerts for staff to review.

The system can detect more than 100 types of issues, including illegal street vending, exposed waste, improperly parked non-motorized vehicles, damaged road surfaces and illegally piled materials. The larger patrol vehicle covers major roads, while a smaller electric vehicle is used in narrow streets and alleys, helping extend AI-enabled monitoring to areas that are difficult for fixed cameras to cover.

“The current recognition accuracy is around 80% to 90%, and we are continuing to optimize the system,” said Gao Huye, an official with the bureau's supervision and performance assessment division. Staff members currently review AI-generated alerts before work orders are automatically assigned to officers responsible for the relevant areas. This helps reduce false positives and duplicate detections while ensuring the accuracy of enforcement.

“Previously, it could take at least 30 minutes from identifying an issue to sending a work order to an officer's mobile device. AI has significantly shortened the process,” said Xie Jinjun, head of the Xinqiao Squadron of the Xinbei District Brigade of Urban Management and Law Enforcement.

Since the pilot program began in the High-Speed Rail New Town area in June, the AI-powered patrol vehicles have generated and helped resolve 258 work orders. Issue detection efficiency has more than doubled compared with traditional manual patrols, while the need for on-street personnel has also been reduced.

The patrol vehicles are part of a broader smart urban management system developed by Xinbei district this year, which combines fixed surveillance cameras with mobile AI-enabled monitoring. While fixed cameras provide continuous coverage of key areas, the patrol vehicles cover blind spots and capture dynamic issues along the streets, creating a more responsive urban management network.

“The system is still in its early stages, so we are improving it through continuous operation, training and optimization,” said Zou Li, head of the bureau's supervision and performance assessment division. As more real-world data are collected and used to train the AI models, the system is expected to become more accurate and adaptable, with the goal of eventually automating the entire work-order process and reducing turnaround time to less than one minute.

扫一扫在手机打开当前页