New research at Constructor University aims to use machine learning for transport logistics

Image
New research at Constructor University aims to use machine learning for transport logistics
Caption
In a new DFG project, Prof. Dr. Yilmaz Uygun is researching how AI can be used in logistics. (Source: Constructor University)

Different formats, frequent updates, more and more data: Managing customer requirements in logistics is a challenge. How can so-called "Large Language Models" help improve this process and make it autonomous, self-learning and with increasing accuracy? This is the focus of a project funded by the German Research Foundation (DFG) at Constructor University Bremen under the direction of Prof. Dr. Yilmaz Uygun.

Customer requirements in logistics relate, for example, to the transport, handling and storage of individual products, their packaging, delivery method or properties. However, there is no general or industry-specific standard for documenting this information. Different templates of very different lengths are in use, which are also updated at ever shorter intervals. Similar problems arise in the processing of specifications, which contain detailed information on complex products and sometimes comprise several thousand pages.

"This is an ever-increasing challenge for the supplier, or contractor, who has to take these requirements into account and implement particular requests," says Uygun, Professor of Logistics Engineering, Technologies and Processes at Constructor University. With the help of machine learning, and particularly large language models, which, like ChatGPT, transfer texts into speech using artificial intelligence, these processes should be streamlined in the future. However, there is still a lack of automated solutions for requirements management in the manufacturing industry.

A problem his new research project aims to tackle. The project, funded by the the “Deutsche Forschungsgemeinschaft” (DFG, German Research Foundation), aims to accelerate the process of addressing requests, improve documentation, and anticipate adjustments needed in the future. The project is set to run for a period of two years, in which Uygun hopes to automate information flow between the supplier and the carrier.

Questions answered by:
Prof. Dr. Yilmaz Uygun | Professor for Logistics Engineering, Technologies and Processes
yuygun@constructor.university | Tel.: +49 421 200-3478

Media Contacts
Name
Rebecca Knecht
Function
Corporate Communications
Email Address
presse@constructor.university
Phone number
+49 175 5861117