Learn2WIn – Optimising Work Processes with the Help of Artificial Intelligence

To ensure that both production and logistics operate faster, the sequence of work steps has to be clear. In Learn2WIn, a research team at the University of Passau is researching how artificial intelligence can help with this. The project is being funded as part of the BMVI project, KIMoNo. 

Whether cooking as a hobby or working in an automotive company, people are confronted with a problem every day, a problem which is being addressed in the Learn2Win project: For things to happen faster, several steps have to run simultaneously. However, which step has to definitively be taken before others can proceed? If artificial intelligence (AI) is to optimise processes in production or logistics, the sequencing of several thousand work steps need to be clear: Should step A be taken before or after step B, or could they run in parallel? This question has to be answered for all the combinations of steps – sometimes this can actually be for several million combinations. In practice, the required data can usually only be collected with the necessary precision by means of expert interviews. However, asking all the relevant questions would take years!

In this Learn2Win project, Professor Otto, the holder of the Chair of Business Administration with a focus on Management Science/Operations and Supply Chain Management at the University of Passau, and her team are trying to develop a kind of dynamic questionnaire which will achieve its aim in an expert interview, with as few answers as possible. An AI instrument will identify a limited number of key questions which can extract as much information as possible, so that a small number of questions can achieve the same as years of interviews with experts. For this to work, after every expert answer, the AI will determine which question should be asked next to obtain as much information content as possible. 

“The approach is very exciting for the automotive industry and logistics specialists,” Professor Otto explains. “Our data collection method enables comprehensive support for process planners using optimisation and AI methods. Thus, in addition to cost-saving potential, other important and difficult-to-handle aspects, such as energy efficiency, ergonomics and learning effects, can be better considered.” 

The project’s acronym, Learn2Win, stands for ‘Learning Precedence Relations with Interviews:  Optimisation Approaches’. The project will run from 1st April 2021 to 30th June 2021. With Learn2Win, the project leader, Professor Otto, had successfully acquired funding in an ideas competition, as part of the BMVI’s KIMoNo project, which aims to investigate and shape mobility in rural areas. Professor Otto holds the Chair of Business Administration with a focus on Management Science/Operations and Supply Chain Management at the University of Passau. Working with her on the project is the ECR, Dr. Benedikt Finnah. The team is also collaborating with Professor Jochen Gönsch (University of Duisburg-Essen).