Tag Archives: Constraints

Finite and Infinite Scheduling in PP/DS

Posted on by

wpid-infinite2-2010-05-3-15-54.jpg

What This Article Covers

  • PP/DS finite scheduling and the its issues
  • Resource maintenance
  • The problem with how many large consulting companies like IBM approach discussing constraints with the client.
  • Being in constraint denial

Constrained or Unconstrained Detailed Scheduling

It’s important to consider and evaluate the use of the sentence

PP/DS performs finite based scheduling.

The reason why it is important to evaluate the comment is for several reasons:

  1. PP/DS can be run in either finite or infinite mode for scheduling.
  2. The value that are entered in the PPMs or PDSs may be “placeholders.”
  3. The constraint values may be incorrect when the project goes live, or may start off as an accurate reflection of factory constraints, and then drift from their initial accuracy due to an inability of maintaining this master data.

Constrained?

I once had a very strange conversation with a consultant who, in a meeting with S&OP, proposed that one way or another the plan would be constrained because the PPMs would have to have values in them for resources in order to “activate.” When I pointed out that these values were not necessarily accurate, he agreed for a moment, but then reverted to his initial track of talking about the constrained plan that would be sent to S&OP.
__________________________________________________

Configuring for Yourself, Not for the Client: How IBM Does Things

I have this experience relatively frequently with consultants from the big consulting firms as they typically like to configure software regardless of its benefit to the company as it is a method for enhancing their resume. This particular consultant worked for IBM, but all the big consulting companies have moved in this direction. With the increasing focus on technical wizardry for its own sake, fewer and fewer SAP consultants understand the businesses they are configuring the system for. It also results in less and less sustainable solutions for clients. I frequently run into mis-designed solutions that were based upon the need for padding billing hours and resume building needs of the consultant by the major consulting companies with increased frequency. You can read more about these issues at the links below.

http://www.scmfocus.com/sapprojectmanagement/2010/04/why-the-big-consulting-firm-implementations-fail-so-often/

http://www.scmfocus.com/sapprojectmanagement/2011/07/the-ibm-systems-implementation-approach/
____________________________________________________

Misunderstood Constraints on a Previous Project

On another project I evaluated their PP/DS installation that they thought was constraining the plan, only to find that the PPM resources were set to infinite capacity and that the only constrain was a lot size.

Finite Scheduling Indicator

Resources can be checked for whether they are constraining, or planning in a finite way by checking the resource master and looking for the finite scheduling indicator. If this indicator is not set, the resources is not constraining. Furthermore, resources can be configured with a “finiteness level,” which means that the resource can be treated as variably finite depending upon the application (i.e. infinite for PP and finitely for DS) Additionally, finite or infinite scheduling can be set in the in global PP/DS settings.

It took over 4 meetings with business representatives who did not believe me, because they said that what I said could not be true because SAP had explicitly told them this is what PP/DS did, and that…

…PP/DS does constraint based planning.

My response was that..

PP/DS has the capability to do constraint based planning, but it must be configured properly to do so.

SAP was effectively hiding this fact from the client, but what the business could not figure out was why their production plan was so unusable. SAP eventually had the analysis into this area buried. The production plan was still ignored by the factory, but the implementation problems were blamed on the business who were “clueless” and just did not “get it.” However, all they had to do is check the SAP documentation, where it clearly states that finite scheduling is only one way of running PP/DS.

Many Problems

There are many reasons for PP/DS implementations to not meet expectations, however, one of the most common is companies have a great deal of difficulty in keeping their resources and PPMs up to date with the actual resource capacities in the factory. SAP does not like to talk about this much, and when they do, they typically see it as a sales opportunity for their ineffective MDM product. Neither SAP nor can other vendors point to many MDM success stories.

http://www.scmfocus.com/supplychainmasterdata/2010/04/why-software-based-mdm-is-a-consulting-boondoggle/

The fact is that MDM will not ride to the rescue on maintaining PPMs and resources, and that the method that SAP has provided to maintenance of this master data is a significant burden on the company to manage. This is why I recommend that companies not get too excited after they take a module live, because the real proof of the maintainability of the systems is several months after go-live.

References

SCM 250 PP/DS Training Manual

Questions? 

Did this article make sense? Was it surprising to you, or have you had the same experiences. If you have any expertise in this area, comment below, and we will respond. 

 

Constraints for SNP

Posted on by

Logical List of Constraints

We found a nice list for helping to appreciate how constraints are managed in SNP. It is from the book Supply Chain Management with APO by Jörg Thomas Dickersbach and sample pages are available at Google Books at

http://books.google.com/books?id=0nJ-nQ2VyXoC&pg=PA186&lpg=PA186&dq=capacity+variants+scm&source=bl&ots=z1ptLCeroN&sig=zTUfSU7unXg0QkUBarCdgsg7FRc&hl=en&ei=DL1LSpfvJ4-MNouogLMK&sa=X&oi=book_result&ct=result&resnum=5

So we don’t think it violates any copyrights to republish it here.

The thing we like about it is it neatly explains some things as constraints, that are not ordinarily thought of as constraints as they are not related to resources (although several of the items listed below are specifically related to resources – as you can see resources apply to the first for items.)

Once you broaden your understanding of constraints beyond simply resources in SCM, it becomes an interesting way of describing the suite.

Contraints for SNP

This does not cover material constraints which can be found here…

http://www.scmfocus.com/sapplanning/2009/09/10/master-data-product-stock-and-product-location-master/

Bottleneck and Finite Resources

Posted on by

Background

SCM has a great variety of resources which can be used to constrain the supply chain and result in a feasible plan. However, the constraint options change depending upon the resource type. In this screen shot below we have a Production Line resource which is used for PP/DS.

Bottleneck Resources

Bottleneck and Finite Resources

This resource can be  declared as finite by selecting the Finite Scheduling checkbox. Checking a resource as a Bottleneck tells the system that the resource in question is the maximum capacity or throughput of the entire process (that is the sequence of all operations) for a product. In PP/DS and production planning generally, the operations are within a single facility. In SNP and supply planning, the sequence of operations may be transportation lanes between facilities, combined with any operations that are performed within the facility (pick, pack, etc..)

What is Necessary to Perform Effective Finite Planning

First, it is only necessary to constrain on a single resource, among a series of predecessor and successor resources. Second, not all resources in the resource chain need to be set to finite. That is because a single resource can be set as the bottleneck resource. Certainly, all resources can be set to finite, (or constrained), but it takes more effort to do this and does not yield advantages because the max throughput is effectively set with a single bottleneck resource. However, whether or not a resource is set as infinite or finite, it must still have a time associated with it.

Critical Products

Products produced on bottleneck resources are called critical. There are two critical resource types generally:

  1. Those produced on a bottleneck resource
  2. Those with long lead times

The bottleneck setting for a resource is on the tab within the resource transaction. It is a simple dialog box.

Critical projects, such as products with long replenishment times or those with constraints are produced at bottleneck resources, which simply means setting the resource to a bottleneck.

To read more about resources in SCM generally see this post.

http://www.scmfocus.com/sapplanning/2008/09/14/resources/

Resources and Constraint Based Planning in SAP SCM

Posted on by


Constraints in SCM

Its hard to overestimate the influence of constraint based planning in the development of SCM. Prior to SCM, or APO’s introduction, constraint based planning was one of the strongest trends in planning generally. A very popular book called the The Goal, which described the theory of constraints, helped popularize an idea that up until that point had been largely restricted to academic circles. For this reason we want to spend some time discussing what they are and how they have become popularized as a mechanism to improve the management of the supply chain.



Constraints Defined

Constraints are simply limitations. A supply chain is filled with constraints. Some of these include:
  • The maximum amount of material that can be placed in a truck
  • The maximum number of miles a truck can drive in a day
  • The maximum output available from a piece of machinery
  • The maximum amount a bin at a warehouse can hold
  • The maximum amount of material a warehouse in total can hold
  • The time windows those are acceptable for material delivery for production facilities, warehouses and retail locations.

Broadly speaking the supply chain has a load placed upon it by demand, which results in procurement loads, production loads, material holding or storage loads and shipment loads. The loads must be serviced by one or another type of capacity.The supply chain can be seen as constantly attempting to manage the highest number of loads with the fewest resources or the highest possible constraints (that is the lowest investment in assets).

Standard Objective

The desire of any company is to maximize loads serviced and minimize the constraints that service the loads (plants, trucks, warehouses, and people).

The concept of constraint based planning migrated from academics to industry only when computerization allowed for the incorporation of limiting factors into software. Previous to computers, these constraints were, and in many cases are still, managed with manually methods. This means people using their experience and judgement to plan and schedule the supply chain.

Resources are one of the most important constraints in SCM. Resources have many control fields which can customize the resource to best emulate the resource being modeled. The resource transaction has many tabs, which allow different types of resources to be configured.

It is one thing to say that a model has constraints, but what is often lost in the statement is how detailed (or should we say available for detailed entry) the constraint capability is within a model. SAP SCM has the ability to enter an incredibly fine level of detail into its constraints. This means that if the customer is willing to put the effort in, they can create a quite realistic and accurate set of resources that can constrain theplan in a realistic fashion. A second issue which is not very often discussed is how capable companies are in maintaining these constraints. The answer is not very. Most companies I have consulted with basically want the system to maintain itself, and the trend towards outsourcing system maintenance in faraway locations where the individuals are purely technical, and lack business knowledge is only exacerbating the situation. One might ask the question if whether the expense of implementations is actually sucking money away from the maintenance of systems. This could be the topic of a separate article by itself.

Finite vs. Infinite Resources

While every resource has the option of being finite or infinite. Finite scheduling means that capacity constraints will be respected. Infinite means they will not be respected, and a second step of capacity leveling is necessary.

Optimization Per Area of SCM

The complexity of optimization is not the same per module. With many resources and more complexity to processes, PP/DS bears the most optimization effort. SNP, which is a higher level of abstraction, is in most cases more straightforward to optimize, and TP/VS, with its few constraints and more explicit costs (transportation, delivery window, etc.) can be seen as the most straightforward to optimize. The fewer constraints inherent to the planning problem, and the more explicitly the costs are (and thus the easier they are to assign a value to), the more straightforward the optimization will be.

Where Are They in SCM? Constraints find their expression in resources in SCM. To read more about resources, see this post.

http://www.scmfocus.com/sapplanning/2008/09/14/resources

Conclusion

SCM began as APO during a different business environment and during the peak of focus on optimization. As this focus has shifted to other areas of supply chain improvement, such as collaboration. SCM has developed from a severely lagging product, to the most widely sold and implemented product in advanced planning. While better than it once was, it still lags in all of its modules. Although SCM does not have a single module that could be considered best of breed, it continues to sell reasonably well because large consulting companies get paid a lot of money in billable hours to recommend it no matter what solution is best for the client. SAP’s development investment in SCM along with its broadening of the role of SCM means that SCM will be more prominent in the architectures of many companies.