Why ‘4.0’ Is Often 3.0, and Smart Is Not Always Intelligent

Why ‘4.0’ Is Often 3.0, and Smart Is Not Always Intelligent
Industry Week June 2021
By Arno Koch
https://www.industryweek.com/operations/continuous-improvement/article/21167832/why-40-is-often-30-and-smart-is-not-always-intelligent

Now that even my fridge has to be “smart” and a simple sensor is called “4.0,” I wonder. Does  everyone have the same idea about the definition of Industry 4.0 and how far are we now really in this process? And what is 4.0 really about? I like to distinguish fact from fiction, and here are my conclusions.

Henry Ford created flow production, which allowed large quantities of the same product to be produced at affordable costs. He described this in a number of (still very readable) books. On the other side of the world, after World War II, these books were closely studied by one of his competitors: Toyota.

Toyota, originally a machine builder of looms, implemented Ford’s principles in detail. However, there were two differences:

1. Whereas in the U.S. everything was abundant, on the Japanese island, with hardly any natural resources, scarcity of everything has traditionally prevailed, giving rise to an incredible focus on eliminating waste.
2. The Japanese have learned that only by working in close harmony with their environment and in cooperation with the people around them, they can produce enough to survive together.

It was precisely these two characteristics that were decisive in the Toyota Production System: By working together in the constant struggle against losses, the famous TPS was developed after World War II during a phase of deadly liquidity problems.

At a certain point, Toyota still had orders and stock, but no more money for new stock. And so it started building only the cars that were already sold with the materials that were in stock—buying what was no longer in stock only for existing orders. And because of good cooperation, standardization and chain thinking, among other things, the now familiar ‘pull’ system emerged, whereby different vehicles are built in a sequence, exactly as the customer requests. This is, in fact, demand-driven mass production based on one-piece flow.

The Japanese also realized that quality and control processes would be essential. And so in postwar Japan, industry quickly adopted Dr. Deming’s statistical process control to provide the basis for flow manufacturing: reliability.

A revolution! Flexible product mix capabilities were built into the process and value chain. Crucially, in this approach, the shop floor is heavily involved in continuous improvement. So, in a way, it was diametrically opposed to what Ford was doing.

The beauty about the pull system is that it works particularly well in extremely complex situations. Western industries started to adopt the one-piece-flow ideas and often added a new control tool to optimize the production process even more: ERP.

The promised idea behind ERP’s material-requirements planning was that it would predict the future—applying  advanced business rules of operations research in the software would provide the maximum result.

In theory, this works perfectly: As long as all parameters are correct and nothing goes wrong in reality, then the calculation is correct, and every part is in the right place at the right time and the puzzle falls together. In practice, a number of uncontrolled situations (quality problems, machine failures, logistic deviations etc.) often give fewer results than expected.

For example, a large aircraft manufacturer had so many uncontrolled situations that it never got a full ERP implementation working, and production only started to flow when it decided to plan the production using a full “pull” system and discard all ERP calculations.

The 4.0 concept builds on this idea of predictability: “If we now put more basic information into the computer and let all the components talk to each other, then we can have more advanced business logic and optimize even further.”

Industry 4.0 was an initiative launched by the German government. It was first promoted at the Hanover Messe in April 2011 by a number of German “influencers” who had every interest in selling more (German!) technology:

1. Mr. Kagermann (ex-CEO SAP)
2. Mr. Luke (Business-Science Research Union).
3. Mr. Wahlster (DFKI: German Research Center for Artificial Intelligence)

The fourth revolution they envisioned was based on two pillars:

1. Each product has a digital twin, where all aspects of the product ánd the production process would be accurately defined.
2. The product, based on its digital twin information, knows where to go to receive what treatment.

To enable this, machines, transport media and product would communicate with each other. It was described glowingly how beautiful it could all be if everything just went automatically. Full automation would be the answer to increase productivity to keep manufacturing in Germany and compete with low-wage manufacturing countries.

According to the German Fraunhofer Institut (2014), this was to result in a productivity increase of 2.2% per year. Two years later (in 2013) the Institut declared, “The third revolution has been successfully completed.”

I have visited several factories in Japan—and also factories from South Africa to the U.S.—that have implemented 3.0, and I have seen the incredible results they achieved. In Japan, I saw zero emission factories, running zero defects for years, where all resources going into the factory came out either in the product or reusable.

Over the last 25 years, my team has implemented overall equipment effectiveness on over 5,000 processes on all continents in any branch. The vast majority of the machines are not capable of running even a day without breakdowns. They stop regularly due to all kind of organizational issues and the products suffer quality losses, often several percent. With this high percentage of uncontrolled situations, I dare to state that the third revolution definitely has not been completed across industry.

Let’s look at this idea closer. The essence of Industry 4.0 is that a digital twin is made of each product: this image contains the complete information about the product, conversion process and production information. The product itself controls the machining, tooling and transport from “conception” via its digital twin: It moves from step to step, where it undergoes the necessary processing until the final product is ready.

To make this possible, machines and transport are connected and communicate with each other and with the product through IoT.

In a 3.0 value stream, the product flows in one fluid motion through the process, from one value-creating step to the next. The knowledge to do this correctly is embedded in the process; through a high degree of standardization and modularity, a high degree of flexibility is obtained. As water is pulled by gravity through the rivers to the sea, in a flow process the products are “pulled” by the customer demand to the customer, hence the “pull” system.

The difference? Whereas in 3.0 we were talking about “value streams” that are fine for a normal person to understand, in the 4.0 concept we see networks emerging where different products move crisscross through each other, all based on the digital twin information. The advantage? We now can make fully personalized products, each configured to the customer’s demand.

Conveniently, first of all this assumes complete and correct digital twin information. And it requires fully controlled processes at all levels.

My question is: Wasn’t this exactly the point of 3.0? If we would make a pair of completely customized shoes for a customer, isn’t that just a normal value-stream, using some new techniques in the value steps, like 3D printing inlays? Where is the advantage of the complex, failure-sensitive digital twin?

Production based on digital twins is now getting extremely dependent on complex technology. As a manufacturer you know the challenges of your ERP system. Am I pessimistic when I say that setting all parameters of a digital twin is not something to look forward to?

Having an IT degree, I am not afraid of new technology. I know about handling complexity. But I also know how solving complex problems with even more complex solutions is a guarantee for even more problems.

The automotive industry has identified this risk—and in the newly released AIAG VDA FMEA reference manual, they have identified a separate risk analysis needs to be made on all software components in a car.

The software issues in the Boeing 737 Max clearly indicate how large the risk is to implement advanced software technology to solve a basic design issue.

If we look across the supply chain, the digital twin is created by the OEM, but the OEM does not have knowledge of how a specific product is being produced at the supplier, nor is the supplier willing to share that proprietary information with the OEM. So how would you be able to create a digital twin in this case?

To realize Industry 4.0’s digital twin concept, a lot of prerequisites need to be available. At the moment, it is safe to say that the vast majority of products and services that carry the “4.0” label are nothing more than potential or assumed tools to achieve more flow, and these tools can be used very effectively in in 3.0 environments.