Fraunhofer Institute for Integrated Circuits IIS
AI-based method for route planning brings greater sustainability to transport logistics
One of the big challenges in dealing with climate change is how to reduce greenhouse gas emissions in transport. In commercial freight traffic in particular, there is significant potential for reducing emissions. After all, a substantial proportion of truck journeys are not loaded to optimum capacity. In the KITE project, our researchers are collaborating with project partners Optitool GmbH, BLG Logistics Group AG & Co. KG and Schmahl & Stoepel GmbH to develop a new AI-based method for route planning. Our goal is to reduce empty runs and make transport logistics more sustainable.
To do this, we are combining machine learning algorithms for forecasting freight volumes with mathematical optimization for route planning and using this approach to expand existing solutions for route planning optimization. By contrast, previous methods are generally able to take only fixed orders into account. Yet good transport schedulers often want to improve the route framework even more by acquiring further orders in a targeted manner, offering or accepting trips on freight exchanges, deferring tours or rejecting trips entirely. In the future, the KITE software will help with these four options by giving intelligent recommendations for action. For example, the algorithm could suggest calling up a certain existing customer who probably has demand and whose order suits the tour. The objective is to reduce the empty runs of the companies involved by up to 15 percent.