Abstract: At present,the traditional belt conveyor inspection method mainly relies on manual inspection,which has some problems such as high labor cost,low efficiency,easy to make mistakes and missing inspection. Therefore,an automatic detection method of mine belt conveyor based on mobile robot is proposed. Mobile robot is used to inspect along the belt conveyor, and the belt conveyor is automatically detected by the mounted sensor and camera,and the information including RGB images required by maintenance inspectors is collected. The automatic detection method of mobile robot is divided into four key parts, including determining the detection target and parameters,designing the sensor and data processing system of mobile robot,developing the detection path and strategy,and real-time monitoring and alarm processing. According to this,the automatic detection process of mine belt conveyor is divided into three steps:automatic navigation of mobile robot,automatic detection of mobile robot and data upload. The automatic navigation function and fault detection efficiency of mobile robot are tested in simulated mine production environment. The results show that the proposed method is helpful to detect the belt conveyor state efficiently and accurately,and alarm in time when abnormal conditions are detected,which provides certain technical support for ensuring the safety of mine production.