What’s the Difference Between Discrete and Process Manufacturing?

We talk to a lot of different manufacturers who generally fall into two broad categories: The discrete manufacturers that make things and the process manufacturers that make stuff.

What’s the Difference Between Discrete and Process Manufacturing?

Discrete manufacturers make things that can be counted and itemized and often require assembly on a production line. Examples of discrete manufacturing companies are the Ford Motor Company making trucks and automobiles, Rawlings making baseballs and gloves, Apple making iPhones and computers, Nike making shoes, and all the other companies making all the other widgets and gizmos. These products are ordered, manufactured or assembled, and delivered, based on “how many” the customer needs.

Discrete manufacturers can have a highly complex bill of materials (BOM), which may be parts or raw materials. In addition, most manufacturers have a multi-step assembly process. If a single part or material is missing, the whole production line or process can grind to a halt. Assembly also necessitates ample factory floor space, multiple machines often organized in cells, and usually requires more human input and labor all throughout the process. For this reason, availability is very important. Without all of these things listed, producing a finished product can be very difficult.

Continuous-flow or Process manufacturers make stuff that has to be mixed from a formula or a recipe. Pharmaceutical companies like GlaxoSmithKline, chemical companies like Dow Dupont, or oil companies like Exxon Mobil Corporation are all examples of continuous process manufacturing companies. Most food and beer manufacturers also fit in this category because of the continuous production methods used, even if the end product is a tangible, packaged food item or bottled drink.

Customers order from process manufacturers based on “how much” of something they need.

Despite their differences, many manufacturers have elements of both types of products in their production process. 

Engineer Stephen Tattershall points out that, “If you use these terms most accurately, they will describe departments, machines or operations of manufacturing rather than businesses or industries. In this context, it is clear that assembly, packaging, stamping, molding, machining, and additive manufacturing are discrete manufacturing, while refining, mixing, filtering, reacting and most thermal treatment are process manufacturing.”

While the two modes of manufacturing seem very different, upon comparison, they do share elements in their supply chains and a need for data, and a desire for control over variability. Whether producing shoes and cars or processing bleach or oil, manufacturing analytics and productivity software play a critical role in making sure that production keeps going.

Control and Quality

Continuous manufacturing heavily leverages real-time controls, often in a highly automated manufacturing environment. This automation is often done with an ERP system and allows for precise measurement and monitoring so quality issues and variation within the product are very low or almost non-existent.

While assembly balances production targets with quality counts and scrap, process manufacturers can focus on making sure that their processes are always running, since as long as the inputs are correct (aka the recipe) then the process can be automated and “hands-off”.

Engineer Ravi Prakash summarizes the difference, “In the process industries we don’t learn to do defect-oriented manufacturing, say if you process a little too much dye, you can just put a little more base in and fix it.”

That’s not to say that quality doesn’t matter, but that quality issues usually aren’t caused by a single faulty machine or operator, but a failure of the input at some point in the batch production process. Issues with recipe variation, mixing temperatures, etc. are the things that most often scrap the batch.

Even highly automated industries like automobile manufacturing utilizing robots and “cobots” requires operators because of the complex tasks involved.

Performance and “Flow” For Discrete Manufacturers

In the book the Fundamentals of Flow Manufacturing, authors Gerard Leone, Richard D. Rahn describe the benefits of applying continuous flow principles to discrete manufacturing, namely that practitioners usually see:

• A 90% decrease of work-in-process inventory
• Dramatic reduction of scrap and rework
• Improved quality
• Improved response time to customer orders
• Better use of floor space and asset utilization

More continuous flow is the goal of most manufacturers so they can make more quality products, faster, and more reliably. This performance metric is often defined as capacity and yield (tracking how much they can make and how much of it is a good product).

For the machine shop floor, high availability and performance mean they can deliver orders early. For make-to-stock manufacturers, it means faster product cycles and quicker time to market. For assembly or make-to-order manufacturers, it means hitting customer delivery dates on time and much less rework.

The key to optimizing this flow will be gaining visibility into the data and analytics of a complex process. For manufacturers seeking to improve capacity and throughput (flow) in discrete manufacturing, they’ll need metrics on all the moving parts of the assembly line:

• What’s our availability of machines throughout the line?
• Why is that machine down?
• What’s our capacity and throughput?
• Where’s the bottleneck?
• What’s our scrap rate?
• Are we on track to hit our production target deliver finished product on time?

These questions need to be answered in real-time to ensure consistent and continuous flow.

Automation Brings “Continuous Flow”

The inherent benefits of process manufacturing and its relatively advanced state of automation when compared to other methods, i.e. real-time monitoring, 24/7 production, less requirement for labor, reduced inventory and inventory management costs, high rate of quality due to strict input/output controls, are enough to make discrete manufacturers green with envy. But, many discrete manufacturing companies are equipping their machines on the line with IoT-capable sensors and controls, adding automation in the form of robots or programmable machines, and tracking machine metrics, as a way of bringing a more continuous flow onto the assembly line.

With more and more assembly lines built around robots or “cobots” (i.e. the symbiosis between a robot and an operator), the promise of “lights-out” manufacturing in which plants churn out discrete products 24/7 with little or no human oversight, isn’t just a smart factory dream anymore.

Even so, assembly is a tough thing to do without human intervention. There are many steps, multiple parts, and tools involved, and the requirements of changeover and complexities of quality control are uniquely suited for human operators.

Automation promises great gains in productivity, but any radical retooling towards automation will need to be informed on the analytics gathered today, based on the benchmarks set by existing machines and their operators.

The case for plant-wide analytics in discrete manufacturing is pretty clear. But what about process manufacturers?

The Importance of Analytics for Process Manufacturers

For process manufacturers, the most important analytics might be measurements of pressure or temperature, as opposed to uptime and cycle times. Quality is more likely to be an automated test, say for PH balance or some other sample assay rather than a visual inspection.

Still, the principles remain, anything that can slow, halt or require rework in production will need to be tracked and monitored.

Depending on the type of process manufacturing there’s almost always a need for PLC-based data gathering in addition to the process monitoring, especially if the process requires human input, like a batch dough mixer, for instance, that kicks off the continuous baking process.

Another need for cycle-based analytics in process manufacturing plants is at the end of the line. Almost every product produced by food and beverage manufacturer, Pepsi Co., for example, will eventually end up in a packaging machine for bottling, shrink wrapping, stacking into cartons, boxes, and palettes, etc.

These machines form a critical link in the entire process, and downtime would mean a bottleneck or a pile-up. Any quality issues at this stage could result in costly scrap.

The types of analytics may differ substantially for each mode of production, but all manufacturers stand to benefit from gathering any data from every machine that might affect availability, performance, and quality.