Bachelor's or Master's thesis: AI-supported navigation system for an autonomous river ferry

Task description:

A major problem for the economic efficiency of local public transport lies less in vehicle and operating costs than in personnel costs. In order to increase the attractiveness of local public transport, the frequency of services would have to be increased considerably, especially in urban peripheral areas and in the countryside. The vehicles could then also be smaller. At present, large and largely empty buses with few passengers run every hour.

One solution would be small, autonomous buses. However, autonomous driving with its high safety requirements for transporting passengers continues to pose a major challenge. The autonomous vehicle must be able to safely navigate dense city traffic as well as country roads with fast-moving oncoming traffic. There is still a long way to go. The thesis associated with this topic is that autonomous driving is easier to achieve in terms of safety for a ferry service on a river the size of the Fulda. Compared to the construction of a bridge or a tunnel, a ferry is a much more cost-effective solution. Such ferries would be highly attractive for accessing local recreation areas, hiking and cycling.

An essential component for autonomous operation is determining the exact position and angle of the hull on the river. The obvious solutions are GPS and differential GPS. However, GPS-based methods reach their limits in the accuracy range of a few centimetres required for mooring and casting off maneuvers. Camera-based methods are currently becoming increasingly important for application sectors such as indoor-outdoor positioning. A camera with millions of pixels is a precise measuring instrument. The challenge lies in the formulation of algorithms to use the measurement information for this purpose. The proposed work topic involves the development and critical comparison of existing algorithms and their further development where they still have weaknesses. The aim is to realize a functional measurement setup based on cameras and possibly additional sensors, which will then be used to carry out measurement campaigns.

Your requirements:

You are interested in mathematics, especially geometry, and have ideally already worked with rotation matrices. You are interested in the development of algorithms, especially in the field of image processing and artificial intelligence. You have completed a degree in computer science, engineering, natural sciences or mathematics and are now looking for a Bachelor's or Master's thesis. You are motivated by challenging tasks. You have a high degree of self-organization and the ability to structure a complex topic in work packages. Finally, as an engineer you are interested in practical implementation and take the opportunity to contact and collaborate with manufacturers, suppliers and workshops.

We offer:

A scientific environment in which your work is embedded in the context of national and international energy system transformation. Our "Integrated Energy Systems" department develops information technology, analytical and numerical methods for the techno-economic evaluation of new energy infrastructures, particularly at system level. An interdisciplinary Institute for Sustainability has recently been founded at the university, through which this work topic is placed in the context of the Faculties of Electrical Engineering and Computer Science, Economics, Civil and Environmental Engineering, Social Sciences, Mechanical Engineering and others.