Changzhou is accelerating the use of artificial intelligence (AI) in its transport system, rolling out AI-driven applications across public transit and highway operations as part of a broader push to anchor new technologies in real-world urban services.

Public transport has been a key testing ground. In Menghe town, a demand-responsive “cloud bus” service completed its first month of operation in December, replacing several fixed-route services with an on-demand network combining physical and virtual stops.

Since its launch, the service has handled 3,781 ride requests and served 5,574 passengers, according to local transport authorities. Average waiting times fell to about 340 seconds, with the shortest recorded wait at three seconds. The fleet was expanded from six to nine vehicles to meet rising demand, while AI-enabled ride matching cut redundant mileage by more than 3,400 kilometres.

Frequent users included residents such as Ding, a villager in Yinxiang, who used the service 66 times during the trial period.

The system is supported by Changzhou’s “public transport digital brain”, a data platform integrating route planning, passenger flow analysis and intelligent scheduling. Authorities said simulation tools were used to model performance before the service went live.

Since its rollout, the cloud bus model has lifted average daily ridership by about 80 percent, reduced vehicle procurement costs by 72 million yuan ($10 million), and cut more than 20,000 kilometers of inefficient travel, officials said.

AI applications are also being expanded beyond public transport. In late 2024, a toll station in the Wujin Economic Development Zone became the first municipally managed expressway facility in Changzhou to complete a full-site upgrade to a “semi free-flow plus cloud tolling” system.

Under the new model, ETC-equipped vehicles are charged automatically via overhead gantries before reaching the toll exit, allowing them to pass without stopping. Processing time per vehicle has fallen from about 350 milliseconds to under 30 milliseconds, with transaction success rates exceeding 99.9 percent, according to operating data.

For non-ETC vehicles, self-service card issuance and payment robots, supported by AI-based vehicle recognition systems, have reduced average entry and exit processing times by 10 percent and 20 percent, respectively.

A centralized “cloud control booth” now allows staff to manage multiple lanes remotely. The operator said manual intervention in exceptional cases had fallen by around one-third.

In September, the Changzhou Transport Bureau released a three-year plan outlining pilot AI applications across highways, ports, railways, traffic enforcement and public transit, identifying ten initial demonstration scenarios.

Officials said the city aims to form a virtuous cycle linking scenario deployment, technological innovation and industrial development, as AI becomes more deeply embedded in transport operations and public services.

扫一扫在手机打开当前页