This book first explores the application of the adaptive Kalman filter algorithm to the estimation of high speed autonomous underwater vehicle dynamics. The authors investigate the performances of different control schemes, from non-model-based to model-based and adaptive model-based, implemented on a low-inertia underwater vehicle for three-dimensional helical trajectory tracking. Control laws for collision avoidance in three-dimensional environments are introduced, considering scenarios where a vehicle detects arbitrarily shaped and nonconvex obstacles using sensors.