AI-based water monitoring with intelligent robotic boats

A robot boat on the water
In the project "RoboBoatAssist" launched in January, researchers from Freiberg are developing artificial intelligence-based control techniques for robotic boats that enable safe, collision-free navigation even in difficult environments with unmapped obstructions above and below the water.

Until now, robotic boats have been controlled by manual remote control or in autopilot mode, where pre-planned routes are followed by the boat. "Both methods require visual contact from the operator to the boat and are also only suitable for open water areas, where sufficient water depth can be assumed. In many cases, this excludes the use in the environmentally most interesting shore zones, which are often not viewable beforehand and where, for example, fallen trees or shallows present unforeseeable collision risks," explains informatics professor Bernhard Jung from the TU Bergakademie Freiberg.

Intelligent assistance systems for robotic-supported navigation of fresh waters

The research work in the RoboBoatAssist project is aimed at developing intelligent, semi-autonomous control methods that, in interaction with operators and artificial intelligence (AI), enable a safe navigation of robotic boats even in difficult situations, such as in the shore area.

Therefore, the researchers are equipping a swimming robot with complex sensor technology such as cameras, LiDAR (Light Detection and Ranging) and sonar for 3D environmental detection above and below water. This allows the robotic boat to provide the operator a view of the conditions on the ground and support him or her in making decisions about navigation targets or routes, and to assist in critical situations, such as navigating around an unexpected object. For example, a sensor-based, interactive route planning is possible that not only takes account of existing map material but also of the currently recorded environment model and the operator's inputs.

Together with the project RoBiMo (robot-assisted inland waterway monitoring), in future regular measurement campaigns will be organized on inland waters in Saxony, to test and further refine the developed assistance functionalities.

Funding by REACT-EU program supports young researchers

The new funding program of the Saxon State Ministry of Science, Culture and Tourism supports young researchers in pandemic times to gain a professional basis. The goal is to keep them in science and to expand their further qualification for current challenges such as sustainability and digitalization in research projects (source: SMWKT press release).

"RoboBoatAssist" is one of 14 funded projects that will receive an additional funding from the European Social Fund (ESF) as part of the REACT-EU initiative (Recovery Assistance for Cohesion and the Territories of Europe).

The Freiberg project (duration: January 1, 2022 to December 31, 2022) offers four young researchers from the fields of computer science and mathematics the opportunity to gain further qualifications within information and communication technologies with application relevance to environmental science and water management. At the same time, the project provides innovation and growth impulses for the recovery of Saxony's economy in the post-Corona period.

In addition to four young researchers, four professorships from computer science and mathematics (Artificial Intelligence and Databases; Virtual Reality and Multimedia; Software Technology and Robotics; Applied Discrete Mathematics) are also involved in the Freiberg research, which is funded with around 316,000 euros.

Logo EU ESF und REACT-EU

Further information:

To the RoboBoatAssist project

To the robotic study course ofe the TU Bergakademie Freiberg

Fragen beantwortet / Contact: 
Prof. Dr.-Ing. Bernhard Jung, Tel.: +49 (3731) 39 38 37, E-Mail: jung@tu-freiberg.de