A study on low-cost sensor fusion for automatic railway coupler inspection

This paper presents findings of a case study conducted to introduce industrial robots into the inspection process of the electrical connector panel in automatic railway couplers. Detection and localization of smaller components in a highly unstructured and dynamic environment similar to a railway maintenance workshop is a non-trivial task. Therefore, a visual perception solution equipped with depth and RGB sensing hardware is suggested. The proposed solution is robust to environmental noise and precise positioning of the target with respect to the camera is not required. The accuracy of the overall system and robustness of the proposed solution are validated with the aid of well-established methodologies, data found in literature and on the actual coupler in usual working environments.