This research aims to develop a wheeled wall-climbing robot (WCR) capable of obstacle climbing. WCRs are widely used for the inspection and maintenance of tall structures, but face significant challenges in urban environments due to obstacles such as ledges and wall-mounted air conditioning units. Adding obstacle-climbing capability to WCRs will improve their adaptability for real-world deployment in complex urban settings.
The proposed robot is equipped with a double active propeller mechanism, which uses two adjustable propellers so that the thrust force can be directed in the desired direction. The thrust force direction needs to be properly directed for obstacle climbing, which is solved by optimizing a set of thrust force direction in a physical simulation environment using genetic algorithm (GA). The optimized trajectories are validated through both simulation and physical experiments using rectangular obstacles to represent typical urban features.
The results demonstrate that the robot can successfully overcome a 20 cm-high obstacle. Although it is not generalizable for other obstacle heights, it already proves the potential of the robot to overcome obstacles using GA-based optimization for real-world scenarios.
While previous wheeled WCRs have demonstrated only limited obstacle-climbing capabilities, this work demonstrates that a propeller-driven wheeled WCR, optimized using a GA-based model-free approach, can effectively overcome such obstacles.
