Building a Digital Twin of the German Rail Network to Deliver Next-Generation Railway Systems
, Senior Software Developer, DB Systel GmbH
Highly Rated
Designing and implementing a large-scale digital twin that's comprised of thousands of kilometers of railway tracks, manifold infrastructure elements, and their direct neighborhood can be challenging. We'll show how we're tackling this challenge step-by-step, starting from exporting the HD map data we acquired from measurement drives to Omniverse down to the perception tasks inside the digital twin environment. In the end, we generate highly accurate synthetic data from Omniverse for arbitrary scenarios that are close to the ones gathered from real-world train operations, such as our Sensors4Rail project. We'll focus on specific challenges that arose when implementing train physics, such as train derailment or speed adjustment, as well as those we faced when implementing digital humans capable of autonomously walking on a train station.