AI-powered drones to help detect hill fires during Chung Yeung Festival

Hong Kong's Fire Services Department plans to introduce three new AI-assisted drone technologies ahead of the Chung Yeung Festival on Sunday to bolster fire detection and response during tomb-sweeping activities. The advanced technology aims to enhance the speed and precision of fire detection, enabling the department to operate more efficiently.

According to Wong Tak-fai, acting divisional officer for information and communications technology, AI not only accelerates the processing of drone images but also assists staff in continuously analyzing large volumes of images from multiple drones at various locations to prioritize situations requiring immediate attention. With the festival approaching, the department will bolster manpower to monitor high-risk locations, intensify district patrols, and launch public education campaigns in collaboration with the Civil Aid Service and other district organizations.

To support the fire-fighting efforts, the department will deploy 159 firefighters across 12 locations in New Territories North, 10 spots in Kowloon, and six locations in New Territories South on Saturday. An additional 166 firefighters will be deployed on Sunday. Furthermore, the department will introduce a newly acquired remotely operated aerial surveillance system, costing HK$1.2 million, manufactured by mainland Chinese company Geely.

This system features off-road vehicles retrofitted with two drone docking systems and a generator, enabling officers to survey fire hazard areas that are otherwise challenging to reach. The drones in this system can perform automated patrols within a 3km radius and are controlled remotely. Additionally, a real-time artificial intelligence video analysis platform will be utilized to expedite the examination of drone footage and images, generating timely alerts for potential fire hazards or drowning incidents.

This platform can process footage in real-time during a drone flight, eliminating the need for officers to manually transfer files from an SD card in the drone to a computer. Currently, analyzing footage post-flight demands human intervention, which can be time-consuming.

Moreover, the department will deploy two high-payload drones equipped with water-bomb systems for deployment against remote hill fires. Each of these drones, priced at HK$500,000 including batteries, can carry up to 150kg and holds an 80-liter water tank, enabling them to release water bombs swiftly. In departmental trials, while it took firefighters one hour to carry rescue equipment uphill for 5km, the high-payload drones could complete the same task in just 10 minutes.

Wong emphasized that the AI would not replace firefighters, as all operational decisions would still be made by humans. The Fire Services Department is exploring the use of large drones in frontline operations, aiming to shift from mere surveillance and observation to actively delivering firefighting equipment.