Comments on Brightwork Articles on Optimization - Brightwork Research & Analysis

Table of Contents: Select a Link to be Taken to That Section

Executive Summary

This article contains comments from articles on optimization.

Introduction

These comments are in response to the articles on optimization.

Comment #1: John Vanerp

Shaun, Not exactly addressing this post, but I agree completely that SAP has not done enough to enable APO to address simulation. You are limited to making version copies and running simulations in a seperate version, but you must go through the painful data maintenance to change parameters in the version specific master data of another version(costs, capacities etc).

I found out recently that SAP is selling to customers a simulation package(not standard) which enables running of the SNP optimizer using the input logs of the optimizer without contaminating the live version(000). The optimization runs based on the input log without affecting any active operational version. In addition, with the custom add-on you can manipulate the input logs directly in tables(costs, capacities, etc) and then run multiple simulations, no impact to active versions. Why this is not part of standard APO I have no idea! This is one of the huge benefits of optimization with simulation versions. SAP has missed the boat on this one for years…John

John,

That is really interesting. I wonder if the simulation package falls under the “Add-In” category that I keep running into. Either way, SAP seems to follow a strategy of keeping a lot of functionality out of the base APO. I get questions about how various Add-Ins, or whatever they call this simulation package you are describing, and I have to answer “I don’t know, I can’t test it.” It puts the client in the position of having to trust the SAP rep. Then a demo must follow, I don’t think I am a big fan of this approach. Certainly, you have heard of COPT-10 parameters, which is kind of similar in terms of being an additional item. I look at COPT-10 and I say “that should be in the base product.”

Thanks a lot for your comment. I was not aware of what you described and I am sure a lot of readers will appreciate the information.

Comment #2: From John Vanerp

“Shaun,
Yes, the simulation package is an Add-in. Obviously it is one of the developments they did for a specific customer and are now selling as an Add-In when it should be part of standard APO. You can see on the presentation in this link that it was dated 2006. Seems that if it was available then, it should have been merged into the standard product by now. I received this presentation from my current client and I am trying to get them to buy the Add-In as it will be very beneficial to the project and would be great to try it out. I uploaded the presentation to Scribd, there should be no problems as there is no copyright claimed in the presentation and it is from ´06.

https://www.scribd.com/doc/94079634/APO-SNP-Optimizer-Advanced-Simulation”

Actually, John, this topic of SAP Add-Ins had me thinking about Plug-Ins that WordPress uses (the publishing system this site is based in). Plug-Ins are interesting in that they allow any developer to extend the application. I have yet to see the concept of an open environment plugin applied to enterprise software. However, it works quite efficiently for WordPress and there is a vibrant plug-in developer community. Plug-ins make applications more desirable because they extend them but without the need for any application integration. I predict SaaS and the SalesForce.com model will continue to grow.

In fact, the way we currently do things, installing applications on-premises is an implementation method from a different era, which it has become the norm, but which is completely inefficient. Enterprise software could learn a lot from how WordPress and other companies like Apple manage their developer universe.

Comment #3: From John Vanerp

Shaun,

I agree that cost optimization has it’s problems for deployment optimization, but the issue I have had with multiple clients is that they want to achieve a fair share, but at the same time consider multiple constraints in an optimal way. Transportation capacity(multiple modes), storage capacity, handling capacity As we know this cannot be done optimally with a heuristic. Heuristic rules can be devised to consider these constraints locally but not globally. So there lies the problem, which leads to many consultants going with optimization as it can consider multiple constraints. There are 2 basic theoretical scenarios to consider(many more in the real world). Shortage of stock and excess of stock. In the shortage scenario, as you said, an optimizer will send all stock to the first option when all penalty costs are maintained at the same level per product/location. For a product with the same non-delivery costs at different locations the optimizer does not differentiate. So it is up to the client/consultant to determine different non-delivery costs for products and locations that meet the business rules. Of course, within the same cost settings, we arrive at the issue of the optimizer sending all stock to the first seeming-random option. But, assigning different non-delivery costs for priority products and locations will reduce the random selection by the optimizer. Not a perfect solution, but if the costs are well maintained it arrives at a reasonable solution that can take into account multiple constraints. This does require good knowledge of the optimizer by the users and discipline in maintaining the costs, which is rare in most clients. In the Excess stock scenario, the storage costs must be maintained higher in the factory than the distribution center. If the storage costs are equal for all products at all DCs then the optimizer has no way to differentiate between them and it chooses the first location on the list to push stock to. So it is necessary to maintain different storage costs per product/DC to achieve better fair share of excess stock based on the business rules. This is one key point to define with the client, what are the business rules and prioritization for the push of excess stock. Many times the first response of the client is that they want to fair share all products..but then they also want to consider transportation capacity and warehouse capacity. They haven’t considered that with these constraints they need to prioritize certain products and distribution centers in order for the optimizer to arrive at a reasonable/realistic solution to the inputs. With Apo 7.0 it is much easier to achieve the fair share objective(still not perfect), as you can define Maximum stock soft constraints at the DCs. When the optimizer is pushing stock it will only go to the maximum stock of the first DC and then push stock of the product to the next DC. Still not a fair share of the product to all DCs but getting closer within the limits of an optimization model.As a disclaimer…..requires intimate knowledge of the optimizer and how penalty costs work, on the part of the client. As a side note, It is perplexing how many APO customers buy an expensive and sophisticated system and not invest in the people that will use the system. Or bring in new hires that have the intellect to use it.

Great comment John. You demonstrate a lot of the nuances of cost optimization. Please send me a shipping address and I will send you a copy of “Supply Planning with MRP, DRP, and APS Software.” Given your real-world experience in cost optimization, I think you will enjoy the book. I am not sure as to the publication date but am close to getting it out.

Would You Like to Comment and Have it Added to This Thread?

Just provide your comment in the chatbox in the lower left of this screen.