This project was carried out by The AnyLogic Company for PREDIT, French programme of research, experimentation and innovation in land transport. The operator of the railway network of the country wanted to know if rail cargo transportation could compete with auto truck transportation. They wanted to make truck-rail-truck transportation more effective. The specific challenges of the project included:
- Determine how many rail cargo terminals were to be built for the optimal functioning of the system. Their characteristics also had to be defined, including location, utilization, and parameters of shipments going through these terminals.
- Determine how the train management was to be organized. This included the train departure timetable (trains could only depart in certain time slots), and the size of the trains (the trains had to be short enough for stations and terminals to be able to handle them, but long enough to be cost- effective).
The decision was to create an Agent-Based model. Terminals, stations, and even some particular segments of rail network were modeled as agents, while shipments, trains and railcars were the elements passively managed by the agents. This approach was used because main decisions concerning the shipments were made at the stations.
In order to compare the two transportation methods, it was decided to double each shipment and send it by combined truck-rail-truck route and by truck only at the same time. The best price and time of shipment was selected by a virtual expeditor and reflected in a database.
The research indicated that with the current structure of the transportation network and the prices, using only trucks was cheaper and faster than the combined truck-rail-truck system. The reason was mainly because of the size and the structure of the network.
The AnyLogic Company provided the customer with a flexible tool for optimizing the existing railway network structure. The user of this tool would be able to locate terminals, change costs, size of trains, etc. to see how the system would react. With this tool, the customer would get the opportunity to see which conditions had to change so that they could achieve their goals of making the rail cargo transportation more efficient.