Railway Digital Twin
Alireza Ghane
Published
Problem:
The Austrian Railway needed a tool to analyze future infrastructure and operational policies for the next 2 to 10 years. Due to the vast number of random variables, traditional methods would requires exponentially large resources with respect to the number of included random variables.
My Contribution
As Chief R&D Engineer, I designed the data models and the simulation framework. I reduced the complexity by modeling data access and simulation services as unique micro-services, accessible by different agents in the system. The stochastic complexity was resolved using a Bayesian graphical model and intelligent timing scheme for track-exact train simulation. The idea behind the time management scheme was borrowed from electronics simulation software, using a combination of discrete times and trigger mechanisms.
The project was partially supported by grant from Vienna Business Agency. Main customer was Austrian Railway.