Micro AUV Localization for Agile Navigation with Low-cost Acoustic Modems

High-rate self-localization for agile navigation still constitutes a major bottleneck in micro autonomous underwater vehicle (µAUV) research. Hereby, key challenges arise from the system being small-size and low-cost, which excludes common approaches such as relying on sophisticated ring-laser gyroscopes. In this context, we investigate two promising sensor fusion strategies for µAUVs based on low-cost sensors and motion models. We benchmark our findings in experiments against a baseline linear least-square estimator and a GPS reference track.