Making transport logistics more sustainable and reducing empty runs is the aim of the ‘KITE’ research project. To this end, AI is being used in transport to reduce emissions. The researchers in our Supply Chain Services working group are using mathematical optimisation and data-driven forecasting to achieve this.
Forecasting and mathematical optimisation for better route planning
As part of KITE, we researched route planning algorithms to reduce empty journeys and thus emissions. We first used data to forecast the »upselling potential«, i.e. the possibility of acquiring additional transport orders, for each route in the network. We then used route planning to calculate which of these potential additional orders should be specifically acquired in order to improve the route structure. Our process is illustrated in a software demonstrator, which also addresses the specific requirements of full load transport for route planning.
An interactive network visualisation for the strategic improvement of the transport network
Transport networks in full load transport are often unbalanced: this means that significantly more cargo is delivered to some districts than is collected from them. This leads to empty journeys, as there is often no corresponding return load. As part of KITE, we have developed an interactive visualisation solution that visualises these imbalances in the network and displays data-driven suggestions for acquiring suitable return load orders.
Division Supply Chain Services at Fraunhofer IIS