In cooperation with ALP.Lab GmbH, the BERNARD Gruppe has developed a redundant, fully autonomous track monitoring system for the Austrian Federal Railways (ÖBB). Based on artificial intelligence (AI), it is used for the early detection of railway tracks being covered by soil, rocks and earth fall.
Any obstruction on an active railway poses considerable risks to ongoing train services. For example, railroad lines can be susceptible to landslides and mudflows due to the shape of the terrain. A loss of forrest cover also contributes to an increase in the risk of railway tracks being covered by earth.
For the best possible identification of obstacles, two independent systems are combined, which both detect objects in the track area with the help of AI: The BERNARD Gruppe uses its optical sensor BMA (BERNARD Mobility Analyser), which detects and classifies objects covering the track. Our project partner ALP.Lab GmbH uses a LIDAR sensor to detect the objects. As soon as one of the sensors detects an obstacle, the system generates a warning message, which is transmitted to the relevant office at ÖBB together with an image of the obstacle.
Martin Meraner, BERNARD Gruppe
