Development and testing of multisensor strategies for autonomous underwater navigation and mapping

In the underwater domain, guaranteeing accurate navigation for an Autonomous Underwater Vehicle (AUV) is a complex but fundamental task to be achieved. As a matter of fact, only by ensuring a correct AUV localization it is possibleto accomplish surveillance, monitoring, and inspection missions. Firstly, focusing on the attitude estimation filters, a strategy based on Extended Kalman Filter (EKF) and Lie groups theory for AUVs orientation initialization has been developed to face the presence of magnetic disturbances, which makes magnetometer measurements unreliable. The procedure is performed when the vehicle is on the sea surface to replace the magnetometer measurements and to evaluate the offset caused by the unknown disturbances. The strategy has been validated and evaluated with real data acquired during experimental campaigns at sea. Turning to position estimation algorithms, most of the navigation filters for AUVs are based on Bayesian estimators, such as the Kalman Filter (KF), the EKF, or the Unscented Kalman Filter (UKF), and employ different instruments, often including the Doppler Velocity Log (DVL) to perform the localization task. Recently, the use of payload sensors, such as cameras or Forward Looking SONARs (FLSs), in navigation-aiding has arisen as an interesting research field in the attempt to reduce localization error drift. Such sensors, if used simultaneously, can provide multiple observations, which can be combined in a Kalman filtering framework to increase navigation robustness against noise sources. Navigation techniques that employ multiple devices can provide a high improvement in the estimation quality, but they can also cause an increase in terms of computational load. Consequently, strategies representing a trade-off between these two conflicting goals have been investigated. Using an augmented set of devices able to provide navigation information represents an intrinsic boost in redundancy: DVL-denied scenarios, such as very close to the seafloor or other surfaces or when a substantial number of gaseous bubbles is present, could thus be managed. Two different UKF-based frameworks have been implemented and compared: on the one hand, a centralized iterative UKF-based navigation approach and on the other hand, a sensor fusion framework with parallel local UKFs. To the author’s best knowledge, the fusion of inertial, acoustic, and optical data in a UKF algorithm and the application of the presented sensor fusion strategies to AUV navigation is novel. It is necessary to highlight that while the centralized strategy guarantees the best improvements in terms of estimation quality, decentralized strategies provide an increase of robustness against measurement reduction. To overcome the limitation of the Kalman filtering strategies and to accomplish the mapping task during vehicle navigation, a factor graph-based Simultaneous Localization And Mapping (SLAM) strategy, which relies on DVL and optical measurements, has been developed and tested. The strategy has been implemented to fuse the data coming from these onboard sensors and to provide a reconstruction of the seabed inspected by the AUV. The proposed solutions have been firstly validated with realistic simulations made through the Unmanned Underwater Vehicle Simulator (UUV Simulator), where a dynamic model of FeelHippo AUV was implemented. Moreover, both the UKF and the SLAM strategies have been tested with real experimental data acquired during several experimental campaigns at sea. Operating on both the orientation and the position estimation filters, the presented works propose advances to improve underwater vehicles’ localization and mapping capabilities.