5G Smart Traffic AI Innovation Technology Inspection and Integration R&D Project
This project addresses the needs and challenges of Kaohsiung Metro Corporation by applying key AI technologies, including AI image recognition, tunnel inspection without GPS signals, unique lighting surface technology, 5G image transmission for bridge inspections, and service robot monitoring. The AI innovation technology inspection integration system is being implemented in selected Kaohsiung Metro locations, including the tunnel between R4A Caoya Station and R4 Kaohsiung International Airport Station, the platform and concourse levels of R15 Ecological Park Station, and the elevated section opposite Exit 1 of R22 Qingpu Station. Inspections are conducted using drones for tunnels and bridges and service robots for station interiors.
Infrastructure Inspection
Tunnel Inspection:
Technology Application: Drones collect spatial and image big data within tunnels. Using AI analysis, inspections covering 1.9 kilometers can be completed in 30 minutes, replacing traditional manual inspections.
AI Analysis: Achieves over 90% accuracy in identifying issues such as segment detachment, water leakage, or foreign object intrusion, improving service efficiency by more than 10%.
Bridge Inspection:
System Functionality: A big data database is established where the drone inspection system automatically categorizes bridge structures. AI provides rapid crack analysis with an accuracy rate exceeding 90%.
Efficiency Improvement: Personnel only need to focus on a small number of abnormal images for degradation assessment. Structural conditions are quickly understood, and degradation types are pre-filled into result reports, enhancing service efficiency by over 50%.
Station Interior Inspection:
Service Robots: Conduct comprehensive inspections on platform and concourse levels, transmitting real-time images for AI recognition. The system automatically logs follow-up tracking cases, improving service efficiency by over 25%.