Intelligent Agent-Based Locomotive Schedule Simulation and Optimisation of Austrian Freight Traffic

AundO is a research cooperation between the Rail Cargo Group Österreich, Technischen Universität Wien, Alpen-Adria-Universität Klagenfurt, Hex GmbH and dwh GmbH - Simulation Services and Technical Solutions, funded by the FFG program Mobilität der Zukunft. The goal of this research project is the development of a tool that enables a rapid, automated, efficient and intelligent locomotive schedule planning of the Austrian freight transport as well as that counteracts additional expenditure and cost, a negative impact on decarbonisation and long delivery times.

Based on huge historic datasets regarding the Austrian rail freight services, a train delay prediction system is built and track capacities are calculated. Specific optimisation algorithms are developed for the planning of the locomotive schedule and routing. Their solution candidates are evaluated by an agent-based simulation (which uses the train delay prediction system) according to their robustness. The simulation results are going to be fed back into the optimisation model via a feedback loop, enabling an effect analysis of the iterative composition of optimisation and simulation.