One of our research focuses has been on the mathematical aspects of sailing. Our group is in particular interested in topics related to Discrete Mathematics, Convexity and Machine Learning. We are working for instance on theory and applications of polar performance diagrams and weather routing algorithms. We are also working on practical aspects of machine learning and created parts of a small "land sailing robot". In a joint Bachelor project with electrical engineers Ulrich Seemann has for instance explored possibilities to collect wind data using a DIY aneometer. Together with velocity data from GPS and an accelerometer one can collect the necessary data for a polar performance diagram. For NMEA-data collection we have configured and programmed with different ESP32s and RasPIs.

We have in particular created a data processing framework for polar performance diagrams and implemented a corresponding Python package hrosailing. It includes a detailed preprocessing framework for the generation of polar performance diagrams. Our software has by now been used in other, quite different research projects and based on our package there exists a specific polar viewer for macOS by Patrick Renner. Since August 2023 there is also a “commercial competitor” AI Polars (by PredictWind) for our open source software.

We also work on Sailing League Problems, that is, on optimized schedules (paring lists) for sailing competitions. We have in particular found an optimized schedule for the national German Sailing League. Some of these scheduling problems are amazingly difficult! For instance, it is still unknown how an optimized pairing list for the national Polish Sailing League would look like. Details can be found in the Bachelor Thesis of Cedric Rönnfeld. Whoever finds a schedule with “fairness deviation” of 2 or 3 for the Polish Sailing League can win a bottle of good whiskey – just let us know!