There are many software products for truck dispatching. Haeger & Schmidt Logistics (HSL) was unable to identify a suitable tool on the market for the digitalized planning of intermodal trains. Therefore the company has developed its own digital solution, the “Train-Optimizer“, which is currently in the pilot phase.
Putting together an intermodal transport by rail is a complex task. Even a single HSL block train can hold 92 TEUs (20ft). With daily departures between the seaports and the hinterland, this quickly adds up to more than 700 slots per week that need to be scheduled. “One of the challenges here is the frequent changing planning data, due for example to delayed seagoing vessel arrivals or departures. Another important influencing factor is the various deepsea terminals to be served. As part of our liner services, we serve five different terminals in Rotterdam alone,” reports Ed van de Velde, who is in charge of seaport planning for intermodal services at HSL’s Ridderkerk site in the Netherlands.
Maximum five minutes for basic planning
In the future, the digital Train Optimizer will support van de Velde and his team in their work. This automatically determines the optimal train assignment based on the existing planning data and restrictions. “The calculation takes a maximum of five minutes, and we run it twice a day,” says Torben Radtke, the brain behind the project. The HSL employee estimates that about 80 percent of the planning can be covered with the algorithm. For van de Velde, this will eliminate the need to manually readjust several times a day when transport-related data changes. He has more time to concentrate on the core of his work: optimizing the processes as a whole, and adjusting influencing factors such as container prioritization if necessary.
“Customers also benefit from digital dispatching. They can already be sure at the time of booking that HSL will identify the best possible connection for the booked containers and adjust it as soon as ocean vessel arrivals and departures shift. This reliability is an important plus for shippers, especially in view of the ever changing supply chains,” explains Jens Möller, General Manager of HSL’s Intermodal Division.
Master’s thesis was the trigger
The starting point for the Train Optimizer was Radtke’s master’s thesis. The original idea was to develop an intelligent algorithm to minimize truck transfers in the port of Rotterdam. It quickly became apparent that the constantly changing planning data, also triggered by current disruptions in the supply chain and delayed ships, were the bigger challenge. The idea of the Train Optimizer, which takes all influencing factors into account, was born.
Best combination of trillions of possibilities
René Kohlhause from HSL’s IT team realized the technical implementation of the project together with IT development partner COCUS from Düsseldorf. The algorithm incorporates all planning data such as cargo opening times and cargo closing times of the ships, the terminals as well as restrictions such as specific mixed loading and weight limits. What then happens in the background is almost incomprehensible to a human being. Ammon Sillah of COCUS clarifies, “Based on the containers to be loaded and the various voyages, there are several trillion possibilities from which the Solver determines the optimal train loading.”
After the calculation, the output is a user-friendly Excel document. “The structure corresponds to the manifest used so far, so that the dispatcher can easily orient himself,” explains Radtke. Currently, van de Velde takes a two-track approach to putting together the train. On his desk, the planning of the Train Optimizer and his manually created planning are on equal footing. By comparing the two, he ensures that all the relevant information has been incorporated into the digitized version.
The Train Optimizer is being used in the pilot phase for the established Alsace-Rotterdam Express. This intermodal scheduled service comprises eight weekly block trains between Strasbourg/Kehl and Rotterdam. Möller outlines further plans: “After a successful pilot phase, the tool will be firmly integrated into the planning process and also extended to other HSL products.”
Also suitable as a barge optimizer
The medium-term goal, he says, is to extend digital scheduling to the regular inland container barge lines that HSL maintains on the Rhine. To make the optimizer “suitable for shipping,” a separate expansion stage would be required. With a slot capacity of 300 TEU per voyage, additional planning data and other restrictions, even more information will be collected. In perspective, the train optimizer can also support with his artificial intelligence (AI) tool. “It is conceivable that in the future the tool will independently rebook containers on other trains or correct obvious booking errors. The prerequisite for implementing AI is that sufficient experience data sets are available,” Radtke classifies.
Already at the start: The Seavessel Optimizer
The Train Optimizer is already the second digitalization tool that HSL has developed itself for its intermodal business. Only recently, the “Seavessel Optimizer” was launched. Using this IT platform, HSL dispatchers know at all times which ocean-going vessels are arriving and departing in Antwerp and Rotterdam and when the corresponding time windows are open for the delivery/collection of containers. Both tools complete each other when it comes to determining the best train slot for the container.