The Mismatch of PP/DS with Process Industry and Yield Variability

Executive Summary

• PP/DS is a significant mismatch with process industry manufacturing.
• This topic is necessary to understand the limitations of PP/DS for scheduling, and we provide specific examples.
• Process industry production scheduling means the ability to handle leading and lagging operations.

Introduction

PP/DS has had few successes in process industry manufacturing. In this article, we will dive into why this is the case.

Experiences with PP/DS in Process Industry Manufacturing

Upon arriving at one client, I observed significant mismatches between the PP/DS design and the process requirements and manufacturing environment. Furthermore, any company which attempts to use PP/DS for process industry manufacturing will run into these same limitations.

However, PP/DS is not the only production planning, and scheduling application inappropriately applied to the industry. Software vendors routinely sell software that can only perform discrete manufacturing planning and scheduling into process industry accounts. Because of the shortage of appropriate applications, process industry manufacturing has one of the highest rates of customized production planning and scheduling applications of any industry.

The Problem Clients Will Face with PP/DS for Scheduling

Before my arrival on some PP/DS projects, most of the project assumed that the PP/DS solution was generally right but that it merely had a few defects. They never knew because their implementation partners never told them that no one has ever been able to get the PP/DS optimizer to work and that the heuristics are so high in overhead that they are also not used. Therefore PP/DS ends up being merely a visual planning board where blocks are moved around manually.

My analysis will demonstrate that the overall PP/DS solution is misdesigned and must be redesigned to include a third-party application for scheduling. As bad as it is for discrete manufacturing, it is far worse for process industry manufacturing environments.

Understanding the Limitations of PP/DS

PP/DS was designed for a minimal set of environments in production scheduling. Simple manufacturing environments work best for PP/DS and companies that want more than what SAP PP offers. However, PP/DS is not close to several best-of-breed applications in scheduling capability.

Many people who are not familiar with multiple scheduling applications may find this surprising, but no one I know who works across manufacturing applications is surprised by this. The problem is that process manufacturing scheduling requires sophisticated and flexible scheduling solutions. Furthermore, food production brings extra complexities, as observed by an article by PlanetTogether, a production planning and scheduling vendor:

“For example there are spoilage concerns, cross contamination must be avoided, cleanouts can be lengthy, intermediate storage is limited and conveyance equipment constrains material flow paths. Production equipment in Food Processing is extremely expensive and therefore limited. Often the output rate and expected yield for a single product can vary depending upon the specific piece of equipment used. For example, one mixer may produce 250 gallons per batch while another produces 1,000 gallons per batch so the time to product x gallons varies based on the mixer chosen.” – Using APS.net for Effective Scheduling in the Food Processing Industry

PP/DS Process Industry Functionality Gaps

The most substantial gaps with the PP/DS scheduling functionality/capability for process clients are the following:

• The ability to specify a changeover time between two different production runs (functionality exists but is problematic)
• The ability to make changes to the production schedule
• The ability to update and manage resources
• The user interface and overall scheduling transparency

The most significant gaps concerning sustainability for this company and PP/DS for scheduling are:

• Too much solution complexity for the company to manage given the resources the company can dedicate to the task
• Excessive master data and overall system maintenance

Variability of Yield/Output

One issue PP/DS has difficulty with is the issue of the process industry variability in production. Variability in output and dramatically complicates the planning process.

For example, in planning for a milk processing company, the following factors affect milk output:

• The amount that it has rained
• The time of year
• The nature of the grass grazed upon

A host of other factors impacts the output product. The output product is the percentage of production that is derived from a set input. The milk, whey, curd, and pharmaceutical fillers are produced for a given quantity of input.

Let us take the example of the planned production for a particular day if the production plan called for 4,000 pounds of whey produced at a plant for a given day. Depending upon the factors listed above. The plant may or may not have the raw material to produce the quantity specified. Instead, the yield could be higher in another product.

Quotation on Yield Output Variability

This is explained in the following quotes.

“In a discrete environment, a product is either good or bad. Each part has certain specifications and if the part meets these, then it is acceptable. In the process industry, it is more complicated. When a lot is produced it has its own unique characteristics. In the chemical industry this might be melt index or the moisture content. In semiconductors, it might be the speed, which the chip will tolerate. Typically the process can be controlled so that these characteristics fall within a range. The lot might be acceptable for some customers, but not for others.

“This semiconductor wafer began as an ingot of pure silicon. It was cut into a wafer, polished, had chemicals applied to it, was etched, patterned with lithography, and had its electrical properties manipulated as part of the manufacturing process. However, as advanced as the semiconductor manufacturing process is, there will still be variability in these chips. Chips that come off of the same wafer, that are right next to each other before they are cut from the wafer, will still have different properties. These wafers will be placed into different bins based upon testing, and will be sold as different products.” –Supply Chain Panning for the Process Industry

Problems with Lagging and Overlapping Operations

An absolute requirement for process manufacturing is the ability to lag and overlap operations. A second operation needs to start after a certain percentage of the first operation is complete. However, PP/DS does not speak in these terms. It understands scheduling two operations simultaneously (SS) or beginning an operation after the first operation is completed (FS). Just this limitation should restrict the application from being selected for process manufacturing, yet it doesn’t.

Inflexible Resource Capacities

Resource capacities must be changed frequently in process manufacturing. However, PP/DS’s PPM, PDS, and Resources only allow for copies to be made and select these alternate objects. However, this creates a significant maintenance issue that many companies are not aware of before implementing PP/DS. In actuality, there has never been a good reason for PPMs and PDS’s existence in the first place. Other scheduling applications keep the BOM, resources, and routing separate and relate them to one another. This allows any object to be changed and based upon links or relationships to be updated. SAP has been actively pushing PDSs over the past several years, but they are even higher maintenance than PPMs. SAP never communicated this to companies when they promoted them.

Dealing with a Bad Design

Overall, PPMs and PDSs are just a bad design that ends up taking up a considerable amount of time and money to support and take up too much implementation time, setting up and even discussing projects. Other supply and production planning applications don’t have this overhead, and it merely something that one must deal with.

The Mismatch

I will say explicitly, PP/DS has little capability to have implementation success in a process manufacturing environment without significant customization and or blending with a best-of-breed solution. PP/DS lacks the functionality the process manufacturing companies require across the board, and the gaps are not even close. The list of incompatibilities with process manufacturing as follows:

These are serious shortcomings. However, this does not prevent SAP or large consulting firms from recommending PP/DS for process accounts.

Where Does PP/DS Apply?

PP/DS might work for companies that have the following characteristics in their scheduling environment.

1. Discrete manufacturing
2. An environment with few changes to the production schedule
3. Very static resource capacities
4. Very sequential start and stop times between its predecessor and successor operations

How to Understand The Poor State of Many PP/DS Projects

This is an example of one of the requirements that come in my email quite a bit. When taken in the context of many other similar project role descriptions, this short role description says some depressing things about the state of PP/DS projects.

This has also been supported by my project experience on PP/DS projects. It’s the type of information you won’t hear from SAP or any consulting companies that implement SAP.

CONTRACT OPPORTUNITY: SAP APO CONSULTANT

-PPDS

-ideal PPDS CIF

-Functional Consulting (no ABAP)

-Going on the system – if bugs, why bugs?

-Customizing

-Support 2/3 level

• From the list above, the CIF is the integration component between APO and SAP ERP. This role description means that long after the application is live, this client continues to have integration problems. Problems with the CIF are a common issue across APO modules, as is discussed in this article.
• Notice below that bullet. It brings up bugs. The client needs someone to debug the app, and PP/DS has a lot of bugs.
• Notice below that is a requirement for customization. In this case, PP/DS cannot meet the client’s business requirements — so long after go-live — further customization is necessary. There are quite a few PP/DS projects like this. Both SAP and consulting companies tell clients that PP/DS can be successful in any environment when, in fact, it has a far more limited area within which it can be successful.

None of these requirements talk about getting more out of the application, user training, advanced users of the application — it is all just getting the application working. Is this really where we are at with PP/DS after all of these years?

The Commonality of These Roles

I see far more of these types of role descriptions than in the past. There are now a few new APO implementations. Most of the APO work for consultants is in fixing systems that are adding little business value. The executives need to see improvement because they made the purchase. And it is not merely for PP/DS. Many of the other modules have similar types of role descriptions as well.

This raises an interesting question about whether this means that the SAP APO’s market share declines in the coming years. One thing is sure; many of the promises that were made about the business benefits of many SAP APO modules have not come true.

This is why I recommend diagnostics be performed on most of the implementations. This is to determine what is working and what is not working –  to make the necessary changes.

Neither SAP nor can any consulting company can perform this type of work. They can’t tell their clients the truth about any topic. This is an analysis of an entirely misleading presentation made by Plan4Demand. Plan4Demand is just like any other consulting company (Deloitte, IBM, etc..) that will tell any lie to get some business. When all you think about is “selling selling selling,” no real testing can occur – and the scientific method goes out the window.

Of course, there are PP/DS implementations that are live and are not looking for consulting assistance. How well these applications are operating is trying to say. That is unless one has the opportunity to go in and benchmark the performance.

Who Keeps Recommending PPDS for Production Scheduling?

SAP and large consulting firms from all over the world are giving companies bad advice regarding the PPDS solution’s applicability to their needs. For years, consultants who want to continue billing hours for implementing a solution, which will never work for their client, have misled companies. I often arrive at these companies after spending large sums of money going down the wrong path.

In previous articles, I have described the problems with PPDS for production scheduling. There are some areas in which any company must be cognizant of PPDS’s limitations; however, the two principal ones are the detailed scheduling interface and the setup matrix or setup table. You can read more about both of these problem areas in PPDS at this link and this link:

For Which Companies Do PPDS Work?

PPDS is composed of two software areas: production planning and detailed scheduling. PPDS has problems with implementations where the manufacturing process is complex and the scheduling operations require anything from average to above-average transparency and flexibility. Transparency is the ability to see and understand the schedule. Flexibility can change the production schedule and have these changes roll throughout the schedule and affect other operations. PPDS is most appropriate for discrete manufacturing operations with a deficient number of changes to the schedule.

Once you get into the details of most companies’ production operations, it becomes apparent that very few environments fit this description. I estimate that PP/DS, and PP/DS, without help from other applications, cannot meet the needs of more than 25% of the environments where it is implemented.

The Approved PP/DS Solution Designs by SAP

There are several different solution designs recommended by SAP for PPDS. They are listed below:

1. SNP to PPDS for production planning and for detailed scheduling
2. DP to PPDS (where a supply planning tool is not used, and the demand plan is created for production orders)
3. SAP ERP to PPDS for both production planning and detailed scheduling
4. SAP ERP to PPDS for detailed scheduling only
5. Production orders are set up by a third-party system and then sent to PPDS for scheduling.

Of the designs listed above, the only ones that work very well have PPDS doing both production planning and detailed scheduling (options 1, 2, and 3). SAP says that PPDS is such a good solution that it makes sense to use it just for scheduling.

After spending enough time using the application and seeing it at different locations, I no longer think this is an advisable design. The particular scheduling portion of PPDS is not a production-ready product, and if such a small portion of PPDS is going to be used, I do not think bringing up PPDS is a cost-effective solution. For many companies that have already implemented PP/DS and have failed because of its scheduling weaknesses, it’s essential to be able to blend the production planning side of PP/DS with a best of breed scheduling application, substantially cutting the DS off of PPDS and making it PP-Best of Breed. This is the only thing that companies (who have anything but elementary scheduling needs) can do to recover their PP/DS implementation.

Who is Continually Recommending PPDS? For The Wrong Environments?

The usual suspects. I continually see large consulting firms like Deloitte, Accenture, and IBM recommends PP/DS for environments for which it is inappropriate.

Part of the reason is that many software selections are performed by the strategy groups in these consulting firms that have never touched the recommended systems.

Secondly, large consulting firms can bill the most number of hours implementing SAP and Oracle solutions. This is part of a general problem in that the consulting companies’ interests are put before the client’s interests.

The outcome is that the percentage of enterprise projects that go-live is brought down. But because the economic structure of consulting is monopolistic, regardless of the track record of or success ratio of previous clients, the large consulting companies continue to get business. For this reason, there is close to zero accountability in consulting, both internally and externally.

• That is, there is no accountability imposed upon consulting companies by clients.
• There is no accountability within consulting companies. According to the large consulting firms, no partner has ever made a mistake.

Partners and their subordinates are never rated on the value they add to clients. The on the internal measurements that the firm has are revenue, recruiting. So things that grow the business and or make it more powerful in the marketplace.

For these reasons, PP/DS’s weaknesses are hidden from clients. A test for appropriateness of the application to the business environment is never actually performed. These types of activities are being repeated across the country. I can say this because a big part of the work that I do helps to recover problem SAP APO implementations. I see the same thing over and over.

Every time I bring up the limited applicability of PP/DS, it is always the first time anyone has heard of it at the client.

Scheduling the Only Solution

For a scheduling-only solution, PP/DS contains too much overhead. It has an optimizer that is difficult to set up and troubleshoot. It lacks critical manual tools such as drag-and-drop rescheduling.

While it has many heuristics, the master data adjustments required to test these heuristics are cumbersome. Thus dialing-in either the heuristics or the optimizer takes a long time. As I have noted above, the scheduling side of PP/DS is so weak that it should not be implemented.

3rd Party Scheduling Solutions

Some third-party scheduling systems are preferable to PP/DS.

• They are reasonably priced.
• They can be installed as either purchased and installed software or as SaaS (software as a service).
• They can also be integrated back to SAP with a reasonable effort because there are not many interfaces between the scheduling system and the upstream system.

Finally, because these applications have a small footprint ( either web-based, hosted, or installed locally), they can be rolled out to multiple factories. Their ease of use supports a shorter implementation timeline. And also a higher likelihood of success over scheduling only PP/DS solution.

Conclusion

PP/DS has failed in virtually every process industry company it has ever attempted to be implemented. And it has failed in most non-process industry companies as well. But it merely lacks the functionality to support process industry manufacturing.

Interestingly, SAP has been promoting how PP/DS is being migrated to S/4HANA. However, the question that should be asked is whether SAP can create a quality production planning and scheduling application. Companies that implemented PP/DS may as well have flushed their license and implementation dollars for PP/DS down a toilet. So why would any company trust SAP’s new version of PP/DS? The highest value for companies with PP/DS is replacing it with a new application or moving back to spreadsheets.

References

*https://www.amazon.com/Supply-Chain-Planning-Process-Industry/dp/0982314817