Background to the research project:

Due to the orientation of production towards concepts such as Industry 4.0, framework conditions of numerous processes for sequence optimization of shop floor production are changing. Here, the sequence can change individually depending on the work order. New production orders, changed priorities at production time, delivery bottlenecks or a changed number of machines require that planning procedures must also consider the environment as an influencing variable. Many different solution approaches exist for the sequence problem, while solution space navigation itself has not yet been investigated. The refinement of an optimization strategy based on the systematic use of dynamically gained knowledge about the solution space is missing.

The research project

An invalid or extremely unfavorable sequence plan may well lead to a very good plan by swapping two work steps. To find such hidden good solutions, methods are suitable which are not computationally located in the solution space. Mutation and recombination can be understood as such navigation strategies. Starting from an initial solution, new points in the solution space are approached. This non-computable search by genetic strategies could be emulated by targeted navigation strategies. The challenge is not only to preserve the character of the unpredictable, but to emulate it ideally or with arbitrary intensity. A review of the state of the art in research shows that while there are a variety of different approaches to solving the
sequence problem, but solution space navigation has not yet been investigated.

Project partners, funding and data

The DFG-funded research project is carried out by the Chair of Information Systems at the University of Potsdam
Funding: 2015-2019
Contact person: Edzard Weber