Design and Development of Machine Vision-Driven Robotics System for Safe and Efficient Inner Pipe Inspection.

This study focuses on developing an inspection robot, integrated with machine vision technology to identify and detect abnormalities on the inner surface of industrial pipelines. This robot can overcome the limitations faced by traditional methods. This robot is powered by Direct current motors on each wheel. To equip the robot with image capturing and object detection capabilities, it was integrated with a machine learning model trained in a Tensor-Flow environment using Python programming. Traditional methods required human presence on site with the implementation of this technology, humans can remotely control the robots functioning via a Bluetooth module and the captured data is displayed on a real-time data monitoring system. With capabilities such as autonomous navigation and advanced image processing techniques, this research aims to develop a robot to enhance safety, efficiency, reduced labor cost, and accessibility of pipe inspection robots for industrial applications. A prototype was developed to demonstrate the stated benefits and tests were performed in a practical environment to ensure robustness and reliability of the system for numerous industrial environments.