Case 1: Solution of traffic problems caused by rapid urbanization in Nanshan District, Shenzhen

Although the urbanization rate in Asia is lower than the global average, it is increasing rapidly as a result of rapid economic development. This is especially true in China, where the urban population has experienced explosive growth through industrialization and urbanization since the reform and opening up. By the end of 2021, China’s urbanization rate had reached 64.7%. Meanwhile, urban road congestion, traffic safety, and a number of other issues have caused widespread concern in China. In order to speed up delivery times, it is not uncommon for take-out riders to violate traffic laws and regulations by driving on the wrong side, running red lights, speeding, cutting in on vehicles, and not wearing helmets. Riders have a large delivery area and are very mobile, while a limited traffic police force makes it difficult to obtain evidence about violations of regulations... In an effort to resolve the above-mentioned problems, local traffic control departments have carried out traffic safety education for riders, increased efforts at remediation, and increased the police presence at key road sections, which have led to an increase in the strain on manpower and material resources. Data from the Chinese Center for Disease Control and Prevention shows that in 2019, electric bicycles accounted for nearly 70% of non-motor vehicle injuries and fatalities in China. On average, one electric bicycle rider dies every hour in a road traffic accident.

As an AI enterprise, SenseTime’s mission is to “be original and lead human progress with AI”, and it is focused on developing an expansive and inclusive AI software platform to promote economic, social, and human development. Faced with the abundance of traffic problems caused by rapid urbanization, SenseTime is determined to provide solutions based on its own computer vision technology.

Nanshan District is a densely populated area with the highest GDP in Shenzhen, and its urban take-out and express delivery industry has grown rapidly. To alleviate the traffic pressure in Nanshan District, SenseTime has developed a set of intelligent vision platforms. By analyzing physical world data in real time with thousands of AI models and converting the data into insights, alarms, and actions, it has provided a multi-scenario, one-stop AI urban governance solution, which realizes the full closed-loop management of AI judgment and disposal and ensures road traffic order and law enforcement management in Nanshan District. As a result, the efficiency of urban traffic in Nanshan District has been significantly improved.

According to actual data, from January to February 2021, there were approximately 800 cases of illegal riders in Nanshan District, a decrease of 56% compared to the approximately 1,800 cases in December of the previous year, showing a clear downward trend. In December 2020, the rider violation record showed the helmet rate to be approximately 44%. From January to February 2021, the figure reached 94%, showing an obvious upward trend.

In December 2020, approximately 140 persons committed multiple consecutive violations, with approximately 40 from January to February 2020, also representing a great reduction in the number of instances. This shows that by meting out punishment, providing education, and promoting traffic safety, rider awareness regarding safe travel has been significantly improved, and the addition of technological means can effectively reduce rider negligence and prevent repeat violations.

As China’s urban development enters a new stage of high-quality development, the model of urban development will undergo major changes, and significant problems such as traffic-related disturbances will become the focus of urban governance. An intelligent traffic governance model based on AI technology will do much to improve the quality of urban development and traffic safety. SenseTime hopes to help improve traffic in more cities. (This case has been recommended and provided by SenseTime)