1/31/2019

Effective production planning is crucial for being competitive, productive, and resource efficient. The allocation of equipment, raw materials, and resources has to be scheduled as exact as possible. Then your plant will process its orders optimally. Your products will be finished as fast and efficient as possible, at the desired date.

So much for the theory! In practise, production planning often works like this: Experienced employees, with the current order situation in their minds, create schedules for the coming days and weeks. This is done either “manually” or by deriving plans from simple models. But this procedure has some serious disadvantages:

  • All in all, generating schedules “manually” is expensive and inefficient. It binds a lof of personal resources for hours (or even days).
  • The method mentioned focuses mainly on conflict-free schedules, not on their optimization. And even when today automatically generated plans are used as templates, it usually takes a lot of rework to customize them for real plants.
  • Production schedules always provide potential for enhancements. But the common method accesses only little of these options. By mathematical optimization, a much bigger search space for better solutions can be covered.

Right here process simulation comes into play! Even today, material flow models generated with INOSIM Software capture the complexities and restrictions of a production plant completely. They provide a Digital Twin of the plant which is exact and dynamical and can be generated fast. In principle, that model can be applied for optimized production planning.

Today, though, advanced calculation methods which enable simulation for scheduling are still missing. INOSIM therefore is planning a research project with partners from industry and science. It aims to develop a software for automatic calculation of schedules which applies evolutionary algorithms. Such approaches can be coupled with any complex model of a production plant, like those provided by INOSIM Software even today. The benefit: Production schedules calculated in such a way can be related directly to the real plant without any rework.

Moreover, the projected production planning system is meant to be reactive concerning plant failures or changed market demands. To achieve this, red-hot methods of Machine Learning and Artifical Intelligence will be used. Hence the software will be able to learn even highly complex interrelationships by using datasets derived from realistic simulations. Reactive scheduling based on real-time calculation will be the result.

The project is expected to achieve an innovative tool for production planning which will offer unique positions to its industrial users. INOSIM and its partners currently work on a detailed project concept. On this website, we will inform you on the development´s further progress,.

Do you still have questions or want more information? Then please contact us.

More news here ...
© 2019 INOSIM Consulting GmbH