Inchworm-inspired micro-inspection robot: Autonomous design for aero-engine inspection

The inspection of aero-engines presents significant challenges due to the complexity of manual maintenance and the limitations of traditional cable-driven robots, which struggle to follow engine rotations and offer limited degrees of freedom and restricted access. To address these issues, this study introduces a five-degree-of-freedom (5DOF) inchworm-inspired cableless robot, equipped with deep learning-based autonomous motion strategies, specifically designed for internal aero-engine inspections. The robot's design draws inspiration from the natural locomotion of inchworms, providing enhanced flexibility and mobility within the confined and complex internal structures of aero-engines. The proposed design includes innovations such as a multi-fold leech-inspired vacuum suction mechanism, which achieves rapid adhesion and release without the need for a vacuum release device. This mechanism increases the suction capacity by 36.5 % and improves adaptability to the curved surfaces of the engine's inner walls. Additionally, a deep learning-based autonomous motion control strategy, utilizing monocular depth estimation and edge detection algorithms, enables the robot to perform autonomous obstacle avoidance and path tracking with a video recognition frame rate of 15 FPS. The robot's performance, including foot adhesion, path-following capabilities, and autonomous obstacle avoidance, was experimentally validated, demonstrating stable and autonomous operation within the aero-engine environment.