Abstract
The demonstrated utility of underwater gliders as measurement platforms for the open ocean has sparked a growing interest in operating them in shallow coastal areas too. Underwater gliders face additional challenges in this environment, such as strong (tidal) currents and high shipping intensity. This work focuses on the probability of losing a glider through a collision with a ship. A ship density map is constructed for the German Bight from observed ship movements derived from automatic identification system data. A simple probability model is developed to convert ship densities into collision probabilities. More realistic—but also more computationally expensive—Monte Carlo simulations were carried out for verification. This model can be used to generate geographic maps showing the probability of glider loss due to collisions. Such maps are useful when planning glider missions. The method developed herein is also applicable to other types of AUVs.