AI drones boost flood response, typhoon observation in China

Heavy rainfall flooded parts of Mile City in southwest China's Yunnan Province on Monday. Local authorities used an AI-enabled drone system to support emergency drainage operations.

As rain continued, authorities deployed a citywide drone network for aerial patrols, providing real-time data for emergency response.

Drone footage guided rescue teams to flooded areas. Rescue teams evacuated stranded residents, directed traffic, opened storm drains and manhole covers to speed up drainage, while personnel remained on site to prevent accidents.

By early Tuesday, emergency drainage work had been largely completed, with normal city life gradually resuming.

In Huai'an, east China's Jiangsu Province, authorities have also deployed a citywide drone inspection system for flood control and emergency response.

After recent severe storms left water levels high across the Lixiahe area, drones have been used for routine levee inspections and water monitoring.

The city has more than 200 drone stations and over 300 manually operated drones, enabling emergency drone deployment within five minutes across the city.

The system also conducts intelligent inspections of 28,000 kilometers of power transmission and distribution lines and more than 190 substations.

As Typhoon Noul approached south China's Guangdong Province, researchers tested swarm drones on Shenzhen's Dapeng Peninsula, conducting hourly observations throughout landfall to improve forecasts.

Four drones equipped with specialized payload sensors flew simultaneously in a coordinated observation network. Different drone models were tested under strong winds and turbulence.

China already operates an integrated weather observation network combining meteorological satellites, radar and ground stations.

However, cloud cover and limited low-altitude observations leave important gaps in measuring the lower atmosphere. High-frequency swarm-drone observations can help bridge that gap.

Researchers said long-term observations in different terrains are still needed to improve typhoon forecasting.