Conceptual design and prototype implementation of a robotic inspection system for sewer geometry analysis

Abstract

Sewer systems are critical components of public infrastructure, requiring regular inspection and maintenance to ensure their functionality and protect both the environment and public health. Current inspections are primarily performed using robotic platforms equipped with pan-tilt-zoom cameras. By adding a high-resolution laser scanner to such platforms, it is possible to capture a point cloud containing geometric information of the sewer system. With a high-performance computation unit, data analysis can be performed during the inspection process to ensure data completeness and quality, detecting pipe segments or damage, and supporting the navigation of the robotic system. This research demonstrates that integrating a laser scanner into mobile robots for underground sewer inspection enables the detection of pipe connections and their alignment. Additionally, it allows for the objective quantification of deformations and deposits, which represent a significant portion of operational defects in sewer systems. Under laboratory conditions, the prototypic robotic system achieved an average point-to-point deviation of less than 10 mm relative to a terrestrial laser scanner reference. By analysing the scanned profiles during the inspection on the robotic platform, accurate sewer models can be derived. By enabling autonomy, inspection capabilities can be enhanced. As more complete and higher quality data are supplied, maintenance strategies, health monitoring, and flood prediction are improved.