With the continuous advancement of urbanization, the use of underground pipelines is gradually increasing. At present, they have been widely used in domestic plumbing supply, oil transportation, natural gas transmission, sewage discharge, etc. However, with the increase of service time, deposits and internal diseases (deformation, cracks, etc.) will appear in the pipeline, which brings potential hidden dangers to the normal transportation of the pipeline. Traditional manual cleaning and inspection is costly and time-consuming, and some liquid and gas are harmful to people’s health, which poses great challenges to pipeline cleaning and disease detection.In response to this phenomenon, we designed and implemented a sewage pipe cleaning robot based on Beidou navigation system, which can help clean the garbage and dirt in the sewage pipe, and can also inspect the diseases inside the pipe. The hardware part of the robot mainly consists of car chassis with a robotic arm, a high-definition camera, a Beidou positioning chip and other parts, while the software part mainly includes an object recognition algorithm based on OpenCV, a robotic-arm grasping control algorithm and a Beidou positioning algorithm. The robot has the functions of grasping foreign objects, accurately positioning, monitoring the internal environment of the pipelines in real time and uploading data to the cloud platform. In the actual pipeline environment, the robot’s functions such as grasping and collecting the foams which treated as foreign objects, real-time positioning and position uploading, and crack detection on the inner wall of the pipeline are tested, and the test results verified the effectiveness of all functions of the robot designed in this paper.