EKF-SLAM for AUV navigation under probabilistic sonar scan-matching

This paper proposes a pose-based algorithm to solve the full Simultaneous Localization And Mapping (SLAM) problem for an Autonomous Underwater Vehicle (AUV), navigating in an unknown and possibly unstructured environment. A probabilistic scan matching technique using range scans gathered from a Mechanical Scanning Imaging Sonar (MSIS) is used together with the robot dead-reckoning displacements. The proposed method utilizes two Extended Kalman Filters (EKFs). The first, estimates the local path traveled by the robot while forming the scan as well as its uncertainty, providing position estimates for correcting the distortions that the vehicle motion produces in the acoustic images. The second is an augmented state EKF that estimates and keeps the registered scans poses. The raw data from the sensors are processed and fused in-line. No priory structural information or initial pose are considered. Also, a method of estimating the uncertainty of the scan matching estimation is provided. The algorithm has been tested on an AUV guided along a 600 m path within a marina environment, showing the viability of the proposed approach.