Joint fault diagnosis of legged robot based on acoustic processing

In legged robot system, certain types of joint faults can lead the entire system unstable since predefined controller cannot function properly after getting damaged. Therefore, the fault diagnosis is an important operation to prevent systems from failures. In this paper, the acoustic-based fault diagnosis for legged robots (AFL) is developed by employing the FFT and fuzzy logic as a feature extraction and a classification, respectively. In the benchmark, the results indicate that the proposed method can detect and inspect the faults of joint efficiently by means of sound, meaning that AFL method is feasible to be utilized and applied in real applications.