Abstract
Robot localization and navigation is one of the fundamental issues involved in robotics. The sewer environment contains only few local features such as manholes, pipe joints and inlets that are usable as landmarks for localization. In order to find an autonomous sewer inspection robot's way to a specified goal (manhole), it has to solve the landmark detection problem. This research describes how a new mobile laser scanner can be employed for robust and fast landmark detection. The methodology is proved with real time laser scanner data from the prototype autonomous sewer inspection robot platform, KANTARO, which carries the camera and other sensors inside the sewer pipe.