How to optimize production order?
The classical approach is the economic order quantity, EOQ. It considers the periodic demand (d), the setup cost (K), and the inventory holding cost (h). We can consider the back-orders, with applying CSL to this model.
However, till now, we have just a certain demand function. When uncertainty is in place (as naturally), EOQ doesn’t work. At least, technically...
The main reason is the complex structure of the total cost function under uncertainty. It doesn’t allow us to find the optimum production quantity, on an easy way.
However, the Unified Supply Model’s cost function provides a good outlook. It may help us to understand the behavior of the function, and figure out the optimum order quantity, (OPQ).
Here, I’ll not enter to the technical details, but I can conclude that; OPQ is very near to improved EOQ. Here’s the formula;
OPQ = Sqrt[ 2Kd / (hCSL)]
Sqrt is square root of the division.
Suppose that the weekly demand is 80 pieces. The setup cost is 200 Euro. The inventory holding cost is 1 Euro/piece. Finally, CSL is 80%.
Then, OPQ => Sqroot[ 2*80*200 /(1*0,8)]=200 pieces...
On this calculation, the assumption is that the demand pattern is “common” type. You may know the normal distribution, and its bell-shape. Like it, if the demand possibility is on a decreasing trend after mean, we can categorize it as “common”.
I know that some demand patterns doesn’t fit it, and can be categorized as “exotic”. For such ones or more precision on “common” ones, I strongly suggest to check the advanced version of Unified Supply Model. It may handle it. You just need a proper coding in a spreadsheet program to apply the algorithm.
And what will be the safety stock?
As I explained in How to optimize inventory, when we optimize the order quantity, it’s very easy to find the safety stock.
You also may check the details of it on Unified Supply Model.
Founder at Arkhon Optima
6 年Dear Catherine, Yes, we can but it doesn't help us for other type of constraints. This article is just about the production order, when there's only a setup cost. However, we know that most of the processes doesn't have the setup, as first priority. Instead of, we have production cycles, MPQ and shelf life, logistics constraints, price leveraging options, warehouse capacity problems etc etc... I would like to continue to add some articles, which will focus to these issues. The unified supply model has a proper solution, or can guide the user. Kindly looking for your comments on these ones...
Founder at Arkhon Optima
6 年Dear Lucio, yes a little bit, because I don't want to enter deep technical discussion. If we have just a setup cost (K), this OPQ formula helps for coming near to the optimum point. Easy to calculate and apply. However, if you need more precision, a detailed version should be in place. If you interested, I can share a review of the technical version, or you may check from the web page. The path is Advanced Version > General Model > Model Definition
Supply Chain-Manager Centralp Développer les compétences pour la performance
6 年Don't know weither we can apply it or not but we need to change the way we approach this EOQ
Supply Chain Professional | Transforming Operations Through Strategy, Technology, and Sustainable Practices | MM at MIT-CTL & BBS Graduate | Certified CBCP, CPIM, CSCP, SCOR-P
6 年A bit nebulous.