Due to the high risks of manual labor in the steel industry, there is a growing demand for robot-based solutions to replace traditional manpower. Steel companies aim to reduce on-site personnel, minimize accidents, and enhance productivity. This study develops a robotic system to monitor conveyors in ironmaking and detect potential bearing failures. Rollers on belt conveyors contain bearings that emit abnormal noise when worn or damaged. Traditional manual inspection requires workers to approach each roller and listen directly, posing safety risks and inefficiencies. The proposed system detects faulty bearings more quickly and accurately by localizing abnormal sounds. The system comprises a manipulator with a microphone on its end-effector. The microphone collects sound along the conveyor as the manipulator moves to detect noise sources. Once an abnormal bearing is located, faster and more accurate maintenance becomes possible. This robotbased inspection method improves safety, inspection speed, and productivity.
