3S Tunnel Defect Inspector

Tunnel defect identification is crucial for tunnel construction. However, it is difficult to detect all defects through conventional tunnel inspection, and inspectors are required to observe with naked eye and manually record inspection results in the busy tunnel area, which is not only cumbersome, time-consuming and inefficient, but it might also lead to traffic incidents due to human fatigue.

• Adopt state-of-art AI and robotics to achieve intelligence. By using deep neural network and transfer learning to replace the human eye detection in conventional tunnel detection, the accuracy can reach millimeter level, and the detection efficiency is also greatly improved.
• The AI chip combined with the edge computing-based neural network makes the AI detection calculation faster and more efficient, enabling engineers to conduct intelligent inspection on-site, while responding to the inspection results quickly and immediately.
• It innovatively combines AprilTag with QR code for positioning, which can continuously and stably detect defects on the surface of the tunnel under different locations and conditions without the need of using GPS technology.

• Inspectors no longer need to use mobile working platforms for inspections in heavy-traffic tunnels, which greatly reduces the need for high-altitude operations in crowded areas, thereby reducing the risk of accidents.
• The detection accuracy is greatly improved to the millimeter level and can cover all parts of the tunnel, helping operators to identify defects efficiently and accurately.
• Following the detection, it automatically generates the digital defect reports for digitalisation of the entire inspection process.
• The successful upgrade and digitalisation of tunnel inspection work will help develop intelligent tunnels. This inspection systemcan also be applied to other large buildings in the long run.

• 2022 Geneva International Exhibition of Inventions - Silver

• Inspection of concrete structure tunnel

• Intelligent inspection drone applied to detect concrete defects in tunnel sites (Under Application)
• Non-GPS In-tunnel Positioning (Under Application)