Guiyang Public Security Traffic Management Bureau (贵阳市公安交通管理局)

Solution

Baidu AI Cloud and Guiyang's traffic-management bureau deployed an AI traffic-signal optimization system for thirty-one key intersections in the city's Guanshanhu core district. The solution combines roadside radar-and-vision integrated sensing devices with Baidu Maps traffic indicators to create a continuous picture of road-network conditions, congestion events, flow levels, and intersection performance. Those inputs feed an AI signal-control large model that generates signal-timing plans for peak, off-peak, and night conditions and updates them dynamically as traffic changes. The system also exposes a monitoring screen that shows macro congestion indicators, citywide five-minute traffic trends, and micro-level alerts for specific intersections or corridors. In practice, the deployment links sensing, diagnosis, timing-plan generation, and operational monitoring in one workflow so traffic managers can move from reactive adjustments to data-driven optimization and earlier intervention.

Data flow

Roadside radar-vision devices and Baidu Maps traffic indicators feed real-time data into the congestion-analysis layer, which diagnoses flow, incidents, and road-network conditions. The traffic-control large model then generates timing schemes for each period and state, and the resulting plans are monitored and refined through dashboards that expose macro, corridor, and single-intersection performance.