"5, 4, 3, 2, 1, launch!" With this command, six drones soared from the Qingdao Metro Dengying depot, marking the official commencement of intelligent drone patrols in Qingdao's railway protection zones (RPZ). This pilot project focuses on the eastern section of Qingdao Metro Line 4, deploying four autonomous drone stations covering the 20-kilometer stretch from Buxi Station to Dahedong Station to evaluate the effectiveness and efficiency of combining autonomous drone stations with artificial intelligence in RPZ surveillance. The goal of this program is to achieve unmanned, automated, and intelligent patrols through autonomous drone stations and a smart low-altitude integrated scheduling platform to explore the effectiveness and efficiency of autonomous drone stations and artificial intelligence technologies in the surveillance of the Qingdao Metro RPZs.

To ensure the safety and integrity of urban rail structures, specific areas surrounding these infrastructures are designated as metro safety control zones known as RPZs which are traditionally patrolled manually, requiring substantial human resources and often leaving blind spots in closed areas or factory courtyards. This method is both inefficient and limited in frequency, presenting a significant challenge for the industry.

Qingdao Metro currently operates eight lines with a total length of 348 kilometers, making regular RPZ inspections crucial. To that end, Qingdao Metro Group has developed an optimization algorithm for urban low-altitude flight patrol routes based on 3D real-world scenarios. Hurdles in key technologies such as real-time AI video recognition and AR fusion of video scenes with spatial information have also been tackled to ensure the effectiveness of this initiative. Drone patrols represent a critical advancement in the intelligent management of RPZs and the autonomous operation of these drones to execute scheduled patrols, identify high-risk machinery and illegal construction activities, and automatically send out alerts would significantly enhance patrol efficiency and address the shortcomings of manual inspections.