Train the train by generating data from a virtual environment

The drive for automation and autonomy in everyday life is on the rise. It is very evident in traffic.

In addition to research in autonomous car transport, rail transport is also booming. To be sufficiently accurate and usable, these autonomous systems are harnessing state-of-the-art approaches to artificial intelligence and machine learning. However, these require vast amounts of data to achieve the necessary reliability.

Very often, part of the data is simulated by creating complete virtual environments, which are often replicas of real environments and also enable the simulation of critical scenarios that cannot otherwise be realized.

IT4Innovations is involved in a project that aims to create parametric virtual environments of a train track and provide a plausible simulation of train sensors.

Petr Strakos
Researcher, developer
IT4Innovations, VSB-TUO