Sensing for Autonomous Navigation Inside Steel Bridges

© 2018 IEEE. The main contribution of this paper is a strategy to build a map of a bridge structure and estimate the precise location of a robot within it. In particular, the focus is on the autonomous navigation of a robot inside the steel arches that support the Sydney Harbour Bridge. A two dimensional laser range finder sensor, rotated about an axis perpendicular to its spin axis is used to capture the geometry of the environment in the form of a set of three-dimensional points; a point cloud. First, the approximate robot location is estimated by exploiting the fact that the environment predominantly consists of planes. Using this location estimate as an initial guess, the iterative closest point (ICP) algorithm is used to align point clouds obtained from nearby locations. Results from the ICP, together with a simultaneous localisation and mapping algorithm is then used to obtain accurate estimates of the locations of all the poses from where information is gathered, as well as a complete map of the environment. Results from experiments are used to demonstrate the effectiveness of proposed techniques