Digital Cognitive Companion for Marine Vessels

The project will evaluate how smart functions can evolve to support unmanned surface vessels and their interaction with human users. The project consists of two main sub-parts:
- Terrain-Aided Navigation: GPS and other GNSS systems are in most cases providing an accurate position estimation. In some situations, they can not be fully trusted, though. In this sub-project, we have used a particle filter to compare bottom depth measurements, magnetic intensity measurements, and bearings to known landmarks with known maps. By this fusion of information sources, we have created a robust method for position estimation.
- Remote supervision of an unmanned surface vessel via Virtual Reality: In this sub-project, we have designed a user interface for remote supervision of an unmanned surface vessel. In a real-world scenario, there is often a limit in the bandwidth between the ship and the user that controls the ship remotely, which prohibits video streams and the usage of high-resolution images. By this, we have used a virtual environment, which has been complemented with small objects detected by the remote vessel. We have conducted a user study, which has shown that the Virtual Reality (VR) approach strengthens the situational awareness capability, and increases the user's ability to detect dangerous situations.