Eye-to-hand visual servoing using 3D stereo laser optics technology for robotic pipe cleaning application

随着北美管道基础设施的老化，为了使该行业可持续发展，对管道的自动化、快速和原位清洗有很大的需求。为此，设计了带有清洁探头的机器人车辆。该行业面临的一个主要挑战是避免机器人车辆在清洁过程中卡在管道中。这种故障是常见的，当沉积材料的深度，由探针去除大于探针执行器可以处理的，由于其有限的允许扭矩。

With aging pipeline infrastructure in North America, there is a major demand for automated, fast and in-situ cleaning of the pipelines to make the industry sustainable. For this purpose, robotic vehicles with cleaning probes have been designed. A major challenge of the industry is to avoid a robotic vehicle to get stuck in the pipe during the cleaning process. Such failures is common when the depth of the deposited material to be removed by the probes is larger than what the probe actuators can handle due to their limited permissible torque. Inuktun Services Ltd. (ISL) is developing high efficiency and robust pipe cleaning vehicles and aims at solving this problem using its vision systems. In this project, ISL and the applicant's research team partner to address this issue. The objectives of the project include development of a method for 3D profiling of the deposit surfaces from 2D camera images; and design of a vision-based technique for positon control of the probes (at the end of robotic arms). Development and integration of the proposed methods will expand the company's market and competitiveness. In addition to the benefits to the company, the proposed work will create training and job opportunities for the graduate students, and will benefit Canada through addressing the growing demand for sustainable pipeline networks for oil/gas/water transmission/distribution.