Experimental Evaluation of a Forward-Looking Sonar-Based System for Acoustic Odometry

Abstract

Autonomous Underwater Vehicles (AUVs) need safe and reliable navigation systems in order to perform the required tasks. Unfortunately, it is well known that the Global Positioning System (GPS) cannot be exploited during underwater missions. As a consequence, the real-time position of the AUV is usually obtained by means of recursive navigation filters such as Kalman Filter (KF), Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF) or dead reckoning techniques where direct speed measurements are usually provided by specialized underwater sensors such as the Doppler Velocity Log (DVL). The final aim of this work is to propose an underwater navigation solution where the AUV is able to safely navigate without the need of a DVL. In particular, linear speed estimations are obtained from a Forward-Looking SONAR (FLS) or, in its absence, exploiting a simple dynamic model. The proposed solution is validated making use of data obtained during test at sea performed in La Spezia (Italy).