A recently completed collaboration between DFDS, DTU, several other universities and companies has focused on how digital tools can be used to make freight transport by trailer smarter and more climate friendly.
You will be familiar with trailers either from transport behind a lorry or from ports, where they are parked waiting to be driven onto a ship that will take over the next part of the transport, typically to another country. Here, the trailers are unloaded from the ship and picked up by a new lorry to transport them to their final destination.
The ships used for transport are called RoRo ships – short for Roll-on/Roll-off – because they have a ramp to drive the cargo on and off. Just as you know from a car ferry, which is also called RoRo.
Mathematical model makes ship loading more efficient
One of the challenges which the collaboration has focused on is the loading of ships. The RoRo mode of transport means that many trailers arrive at the port shortly before departure. When the loading of the ship begins, it is therefore still uncertain how much cargo the ship must be able to carry – and how much the cargo weighs.
"We have investigated how to optimize the loading of RoRo ships under these conditions. It turned out that, with mathematical optimization, we can plan the placement of the cargo on the ship so that the weight distribution is as good as possible to ensure stability and optimize the ship's position on the water. Our study also indicates that the model can be useful even if 60% of the cargo has not arrived when loading begins and we do not know when or if it will arrive," says Professor Dario Pacino, DTU, who has been the leader of the collaborative project.
The researchers have developed a large mathematical model for use during loading. Dario Pacino emphasizes that it is a large model because it must be able to accommodate the many variables and constraints involved in the task. The model ensures that the cargo is distributed so that the weight on the ship is as optimal as possible. This means that the many ballast tanks on the ship filled with water to compensate for any imbalance, are not used to the same extent as before.
"The ballast tanks on a ship like this can hold a lot of water. So, when we use the model to significantly reduce the amount, it has a major impact on fuel consumption and thus climate impact," says Dario Pacino.
During the project, the researchers attempted to use artificial intelligence to assist the planning. It was not straightforward, as many different factors need to be considered in the solution, and much of the data is uncertain or unavailable at the start of loading and only becomes available as the trailers arrive.
"After investigating the possibility, we have concluded that it will be possible to incorporate artificial intelligence into the solution. It means that we will continue our work," says Dario Pacino.
A start-up was born along the way
Another area that the collaborative project has explored is identifying trailers at the port to ensure faster loading.
"Containers have an identification number that is always located in the same place and written in the same way. It is completely standardized. On trailers, the identification number can be located anywhere and written in many ways – perhaps just with chalk," says Head of Innovation & Partnerships Mads Bentzen Billesø, DFDS.
"At the same time, it is a challenge that the trailers are sometimes not located at the parking space assigned to the truck driver upon arrival at the port. If the driver reads the number incorrectly or cannot find the right space, the trailer is placed somewhere else than noted in the system."
A solution for identifying the trailers has been part of the collaboration. The result was so good that the project gave rise to a start-up company, VizTrack. Behind the start-up are researchers from the University of Tallinn, who during the project developed software and a vehicle equipped with cameras that are used to find the trailer's ID number and identify its location in the port.
VizTrack's solution, which provides the desired overview of the trailer's location, has already been marketed to ports around the world.
Industrial PhD to look at artificial intelligence
The effort to investigate how digital solutions can help RoRo transport has also focused on limiting the time RoRo ships spend in port by loading and unloading simultaneously.
"We can use new digital tools to support our work in loading and unloading ships even faster and to optimize the operation of tugmasters, which we use to move trailers at the terminal. It will be a great benefit for us to be able to reduce the time the ship is idle in port," says Mads Bentzen Billesø.
The effort continues with the appointment of an industrial PhD at DFDS, who will further develop the solutions and take the next step in how artificial intelligence can be incorporated into the planning of loading and unloading RoRo ships. The industrial PhD will spend half of the working hours at DTU and the other half at DFDS to link research and practical solutions.
"We have been collaborating with DTU, among others, on projects for many years and would like to continue doing so. This helps us ensure access to the latest knowledge and research, which we can use to develop new solutions," explains Mads Bentzen Billesø.
