Machine learning in practice

Kiel – People, nature, technology - there are interesting possibilities for using artificial intelligence (AI) in every area. How these could look like exactly for Schleswig-Holstein was the subject of the Science Match, to which the State Chancellery invited almost 50 experts from all over the country. In intensive lectures they presented their work and ideas in the field for Schleswig-Holstein.

And so did Dr. Torben Ott, Senior Consultant for Data Science at Consist. In his lecture he discussed interesting application scenarios of machine learning. Image evaluations are a part of this and are becoming increasingly important, especially for the predictive maintenance and preservation of assets.

The advantages of image recognition with regard to the time and cost-optimized maintenance of the infrastructure of the railway were explained using the example of train wagons. The challenge here is to find the optimal maintenance time to minimize downtime. In addition to their observation in operation, so-called deep learning, in this case the in-depth evaluation of photos of passing trains, maintenance requirements can be detected in good time. Certainly not only for Schleswig-Holstein an important cognitive value.

This can certainly be applied to the conservation of natural resources, such as the typical “Knicks” of Schleswig-Holstein. Here, Torben Ott outlined the idea of using information from aerial photography with methods of deep learning in the future. This shows how diverse the possibilities and application scenarios of artificial intelligence are.