Towards Autonomous Ship Hull Inspection using the Bluefin HAUV

Abstract

In this paper we describe our effort to automate ship hull inspection for security applications. Our main contribution is a system that is capable of drift-free self-localization on a ship hull for extended periods of time. Maintaining accurate localization for the duration of a mission is important for navigation and for ensuring full coverage of the area to be inspected. We exclusively use onboard sensors including an imaging sonar to correct for drift in the vehicle’s navigation sensors. We present preliminary results from online experiments on a ship hull. We further describe ongoing work including adding capabilities for change detection by aligning vehicle trajectories of different missions based on a technique recently developed in our lab.