What are the problems with leveling

How to optimize your production leveling

How does production leveling work?

Production leveling is about an even distribution of the production volume (total and per product) over individual periods. A period can be, for example, a shift, a day, a week or a month. The aim is an even utilization of production and a high degree of flexibility with regard to fluctuations in demand. Production leveling is an important part of implementing the Heijunka principle. However, leveling production often also plays a decisive role in long-term capacity / production planning.

When the complexity becomes too great

In theory, the total amount required is simply divided by the number of periods. But in practice this trivial procedure is not possible or sufficient in most cases. As soon as further restrictions and conditions have to be taken into account, a good distribution of the production quantities per product and overall over the individual periods becomes an extremely complex task for which humans are not suitable.

There is no clear strategy with which one can come to a good solution step by step. The number of combinatorial possibilities becomes extremely large very quickly.

Optimization criteria for production leveling

The planning task consists in the allocation of orders to production periods so that the product mix and other optional criteria such as secondary resources and priorities are optimized. The following optimization goals are supported and can be configured appropriately for the specific application:

  • Leveling the total production volume over all periods.
  • Leveling the production volume of each type of product over all periods. This means that a similar product mix is ‚Äč‚Äčachieved in each period.
  • Due dates for orders: Injuries are possible, but they can be made as expensive as you want.
  • Prioritization of orders: Each order is assigned a priority and the algorithm tries to assign orders with higher priority to earlier production periods.
  • Minimum and maximum use of secondary resources per period: Here, too, it is possible to exceed or fall below the limit and the price for this can also be configured.

In addition, you can also have the batch sizes optimized if you wish. This can bring enormous improvements in the balanced distribution, for example if the lots are chosen at random in such a way that a well-balanced distribution is impossible. In this case, the Production Leveling Algorithm will divide or combine the lots in such a way that the above-mentioned optimization goals can be achieved in the best possible way. The lot size optimization, like each of the optimization goals, can be configured optionally and flexibly.

Creation and use of the MCP Production Leveling Algorithm

The algorithms for production leveling were developed and scientifically researched in cooperation with the Vienna University of Technology (Christian Doppler Laboratory). There are various recognized methods of metaheuristic local search, such as simulated annealing, for use. The metaheuristic framework developed by MCP was used for implementation.

The production leveling algorithm is already being used successfully in several plants as part of planning with the APS tool Preactor (now: Opcenter) and in conjunction with EPEI planning. We were able to show that our solution methods also in practice find solutions in a short time that come very close to the optimal ones. Further information on the scientific results can be found here.
The algorithms developed are independent and independent of technology. They can be called directly from Visual Basic and C # right now. The optimized production leveling can also be used as an add-on in Microsoft Excel.

Conclusion

Production leveling is a standard method for process optimization and increasing productivity. It can be easily integrated between lot size planning and detailed planning. Likewise, leveling production can play an important role in long-term capacity planning. Among other things, the product mix and the use of secondary resources can be balanced. In addition, due date violations can be minimized and lot sizes can be optimized.

Are you having problems leveling production?

We would be happy to advise you on how production leveling can also be carried out smoothly and efficiently for you and how production can be balanced and efficient. We would be happy to present the MCP Production Leveling Algorithm to you in a webinar.

About the authors

Christoph Mrkvicka is a partner at MCP and heads the Research & Development department. He studied technical mathematics at the Vienna University of Technology and has been working intensively on the subject of production planning for 7 years.