Autonomous regional trains
Autonomous trains are appealing for different reasons. First, they would allow more train traffic despite a lack of train drivers. But digitization of train operations also helps to reduce costs and it enables better service for customers.
But currently autonomous trains are only possible in a very limited area of application (e.g. underground railways). To enable the use of autonomous trains in regular traffic, a lot of challenges must be solved. One important aspect is object recognition, from the position of the tracks to the detection of obstacles. For this problem, machine learning is a possible approach, but the quality assurance of AI-based systems is a challenge. To make such systems reliable, a lot of work remains to be done.
Text: Elite Graduate Program "Software Engineering"