AI Speeds up Transportation Logistics

As an example, Friedrich cites a gas supplier that has equipped its gas containers with sensors that monitor how full they are. The system uses this sensor technology to check the current status continuously and calculate when the customer will need a new delivery. “With this information, the AI can predict demand very accurately and can tailor warehouse and fleet management decisions to these predictions,” says the scientist. He’s describing the current state of play, because generating predictions based on internal and external data and using them as a basis for making new decisions is already standard technology. “Now, we’re coming to the language model level. For this, we have to combine massive quantities of information with operator communications and process them in such a way that the content is easier to understand and access. We’re hoping to present a prototype of this technology at Fraunhofer IML for the Zukunftskongress conference. This will mainly focus on knowledge management and representation in the logistics context.”

Another tool that is set to trigger a mini-revolution in transportation logistics is the digital twin, a virtual reproduction of a physical object or system that collects and analyzes data in real time. “It’s important to note that a digital twin is more than just a 3D model. It is a dynamic, digital representation of a physical object or system that receives continuous updates from sensors and other data sources. For example, in transportation logistics, a digital twin could map a vehicle, a fleet, a logistics site or an entire logistics process,” says Achim Klukas, a team leader in the Transportation Logistics department at Fraunhofer IML. 

Fraunhofer IML researchers are currently involved in a number of projects that aim to implement digital twins in the field of transportation logistics. For example, project CRISTAL (Climate Resilient and Environmentally Sustainable Transport Infrastructure) is focusing on increasing the share of freight transportation on inland waterways (IWT) by at least 20 percent and demonstrating strategies for improving reliability by 80 percent at its three pilot sites (in Italy, Poland and France). “The project will help ensure that IWT capacity can be used to a greater extent than before, even during extreme weather events,” explains Klukas. To facilitate this, the CRISTAL project team have equipped buoys with sensors so that they can monitor water depth, which will then be used by infrastructure managers in conjunction with their DSS-RIS layer to determine the water level under a ship’s keel. The results the researchers are hoping to attain include ready-to-use guidelines for establishing a synchromodal transportation corridor management system, covering regulation, administration, business models and management. They are also planning to set up a comprehensive real-time monitoring system for water levels and hydrological conditions, including RIS. “The solutions developed in project CRISTAL will strengthen the logistics processes in inland waterway transportation, enabling us to exploit previously untapped potential,” concludes Achim Klukas.

Meanwhile, the TOLKIEN project team aims to increase process transparency at inland ports and to provide recommendations for process optimization. “One key focus in the project is developing a comprehensive, validated database and conducting standardized digital mapping of processes and emissions in the port area,” explains Klukas. Other focus areas include using artificial intelligence to fine-tune recommended measures for selected business cases. The recommendations will concentrate on emission reduction and traffic flow control and will be incorporated into the digital map via machine learning methods. Fine-tuning of these business cases will be conducted in an agile manner, in line with the current project conditions: “The aim of the project is to identify and amalgamate individual data sources at an inland port and to build a uniform, structured database. This will allow us to create a consistent representation of the processes and thus increase visibility of emissions and traffic flows. To achieve this, we aim to create a digital map and model – also known as a digital twin – of the port’s business processes.”