What Makes Process Manufacturing Unique to ERP systems?

An Insider view for selecting an ERP for Manufacturing companies in the Food & Beverage, Chemical, Ink, Pharmaceutical & Nutraceutical Industries.

In the last few years, many business application software vendors have announced a focus on process manufacturing, specifically targeting food, chemical, and pharmaceutical sectors. Their secondary markets often include paper, building materials, and textiles.  At the same time, many of the larger ERP vendors are announcing everything from products to centers of excellence for many other vertical markets—government, retail, healthcare, etc.  One such vendor’s website states that it is “totally focused” on eight different vertical markets, one of which is process. Buried way beneath these advancements lies the truth that most vendors have created a solution that was designed for the discrete industry and have “adapted” them or “bolted on” functionality to cater to specific process industry requirements. [break2] So what makes process manufacturers different? [break2] To understand if you are a candidate for Process ERP, you should look at the reality of your business. Business characteristics that create differences in requirements, which in turn create differences in the software, are the components that you need to understand when looking at ERP.
Looking at various components of ERP, we see that some components such as financials have very few, if any, characteristics that are specific to a vertical. A financials application can be implemented in any vertical with minimal industry – specific issues, whether it is a steel works, a car plant, a bank, or a dairy. We notice the greatest variations in the supply chain in inventory and production issues—those dealing with the processing and utilization of materials to satisfy ever-changing customer demand. Process characteristics encompass basic characteristics of materials, the predictability of both materials and processes, and the shape of the bill of material.  Process materials vary from natural resources such as solids, powder, liquid, or gas on the raw material end to either the same for industrial companies or, in cases of product for consumer goods companies, on the finished product end.
When a discrete manufacturer specifies materials (parts) to his vendor, he gives a part or drawing number, very tight specifications (plus or minus 1/1000 of an inch, for example) and exact delivery dates. In the process world, the materials are defined more broadly, with a wider specification (water content from 3 to 7 percent, for example) and dates that may be broadly specified due to seasonality or transportation uncertainty.
A good software system is based upon a model of the business
In the discrete world, the primary basis of the model is the Bill of Material. One base assumption is that some number of components makes one of the end items. Since Bills are traditionally drawn with the consumed parts at the bottom and the produced part at the top, the Bill of Material has the basic “A” shape. This does not accommodate the many by-products, co-products, recycles, waste, multiple stage processes, product variations, and yield losses/gains you may have in many processes.  It is not that the designer of the Bill of Materials set down a series of incorrect assumptions about the business characteristics that he was trying to address—he just limited his thinking to the needs of a specific kind of manufacturing operation — discrete.
What are the major assumptions inherent in the Bill of Material and what are the process realities?
So how much process are you? Some process companies have relatively simple processes (receive bulk materials, mix and package) and deal with materials that are very consistent. For these “simple process” companies, any ERP system could be considered, including discrete ERP products with added process functionality. But complex process companies must be much more selective. The additional requirements of their environment may not be adequately addressed by “multi-purpose” solutions.  Those vendors who truly specialize in the process industry should offer more process functionality. They simply cannot afford to be too discrete.  Talk to us about how Sage can deliver!
Examples
In production and logistics, the system must allow management of materials based on the materials characteristics that are important to success.  Management means storing inventory and selecting the right materials for customer order or production based on shelf life, quality, location, potency, best before date, and the like, to maximize customer satisfaction or production performance.  One test of a system’s ability to meet specific needs is to check the database for required item characteristics :  if the needed characteristics are not provided as fields, the system cannot likely provide the management function.  The discrete manufacturer gets relatively few surprises. Order 500 of part A from a vendor and he delivers 500, and the materials received either pass or fail to meet the specifications. Part A always fits with Part B and together they always make sub-assembly C. Set out to make 100 of item C and you almost always get 100 of C, or perhaps 98 good ones and 2 in the scrap bin. In the assembly of A and B, you never end up with item D!  Contrast this to the typical process manufacturing company situation.  Specify 1000 pounds of material X with a range of acceptable specifications.  The vendor often tries to deliver 5000 pounds, but the shipment delivered is usually not exactly that amount. Test the material to see if it meets the specifications and record the results.  Additional materials may be required (buffers, dilution, and other compensation ingredients) to adjust the material to make it usable.
Let’s look at another example: a production formula states 100 pounds of A plus 100 pounds of B produces 195 pounds of C.  In reality, each time C is produced, a slightly different amount results (yield loss or gain).  Sometimes, you don’t even get C at all – you do get a product that is close to C (call it D). Sometimes you get some of both.  The result often cannot be predicted very accurately.  A process ERP must be designed to accommodate all of this uncertainty.