The paper describes an autonomous water vehicle (ASV) capable of autonomously mapping shallow water environments above and below the water surface. Over the past two years, Fraunhofer IOSB has developed a system that is fully electrified and equipped with extensive sensor technology (multibeam sonar, lidar, cameras, IMU, GNSS). For autonomous navigation, the complete processing pipeline was implemented, from obstacle detection and avoidance to trajectory planning and control to multi-sensor localization and mapping. Above water, both lidar-based mapping and photogrammetric methods are used; underwater, bathymetry data is obtained using sonar. The interface to the operator is realized by an interactive digital map table, which allows intuitive mission specification and evaluation.

Simulation plays an important role in the development, testing and evaluation of new robotic applications, reducing implementation time, cost and risk. In this paper we show a digital twin simulation model of an inspection ROV which is capable of performing structural health monitoring by automated creation of a map of an offshore wind monopile. The data is compared to a known reference model. The digital twin simulation model is extended by a physical sensor data input device to bridge the gap between simulation and testing in water.

We show a multi-sensor platform for mapping tasks of shallow water areas like rivers or harbors, both under and above the waterline simultaneously. The autonomy of the platform is realized just by integrating the sensors and the autonomy box which provides, i.a., the motion planning. The sensor data is synchronized purely in software by a newly developed timestamping filter. Maps obtained by the autonomously driven platform are shown as an example for the platform capabilities.