This study developed an intelligent building exterior tile inspection robot based on the movement patterns of geckos and inchworms. The robot is equipped with AI visual recognition and AI audio recognition, using a miniature AI camera and a tapping device for defect detection. It utilizes TinyML and CNN technologies for data processing. The system can instantly transmit inspection results and features obstacle avoidance to enhance inspection efficiency.

The innovative features of this study＇s GLEWBOT include: bio-inspired design combining gecko adhesion and woodpecker tapping techniques; real-time audio analysis using TinyML; integration of CNN for defect classification; autonomous defect recognition; modular design and intelligent obstacle avoidance system for improved efficiency and safety. Its innovation has been recognized by platforms such as Circuit Digest, Arduino Blog, and Hackster.io.

This technology＇s industrial potential includes applications in buildings, bridges, offshore wind turbines, and oil pipelines. Its automation, precision, and efficiency can reduce maintenance costs, increase work efficiency, and decrease risks. Widespread use of this innovation will advance inspection industry technology.