If we were to use Stratasys’s share price as a barometer for the potential of 3D printing to redefine manufacturing we now know that 3D printing’s potential is somewhat of a hoax perpetrated on unsuspecting investors by many. The stock is currently 85% off its peak. 3D printing has its place, but it is not (in the short and medium term) going to remake manufacturing and supply chains in most industries. You can read the details in this comprehensive paper by NIST.
This morning I got up and found an email from BCG in which I discovered that this article on Industry 4.0 is one of the most read. The article and its facetious nature ticked me off a little bit and prompted me to write this. The article might as well have been written to see how many buzzwords can be stuffed into a high sounding paper. Worse would be to find that a large number of people are impressed by this paper.
The misconception starts right in the first paragraph.
The steam engine powered factories in the nineteenth century, electrification led to mass production in the early part of the twentieth century, and industry became automated in the 1970s. In the decades that followed, however, industrial technological advancements were only incremental, especially compared with the breakthroughs that transformed IT, mobile communications, and e-commerce.
To the authors: Toyota Production System, one of the most important innovations in manufacturing in recent decades, was developed, fine-tuned and broadly adopted in the 1970s and beyond.
The next paragraph is even more egregious in its use of bloviation and exaggeration.
Now let’s look at the nine foundational technologies selected by BCG that make up Industry 4.0 transformation.
Big Data and Analytics
Statistics arrived on the scene a good hundred years before “Big Data and Analytics”. Manufacturing companies have been using statistics with and without big data for several decades now. I do believe that advanced techniques such as Machine Learning have an increasingly role to play in manufacturing. But Big Data and Analytics is not as ground shaking as the article suggests as far as manufacturing is concerned. These are many fancy use cases around Predictive Quality using Big Data (analyzing all data that is), but more often than not Source Inspection and Poka-yoke strategies are better.
Besides and in many cases a simpler approach is often faster and cheaper than the big data approach as can seen in this post.
This is Elon Musk’s biggest fear. Imagine autonomous robots attain consciousness, communicate and go on strike.
I have no idea what the authors are talking about here. CAD/CAM/CAE and discrete event simulation have been around for decades and have been used extensively in manufacturing domains. What about SMED?
That said simulation of complex systems (integrated mechanical, electronics and communications systems) continues to be devilishly difficult and advances in this area can help design, development and manufacture of intelligent products.
Horizontal and Vertical Integration
Companies have been attempting cross-functional engineering for decades now. It’s been hard to accomplish because of human nature and tools provided by vendors suck.
Vertical integration and its practice is dependent on company’s core competencies, market dynamics and how much control a company wants to have in the value chain.
Industrial Internet of Things
Less said the better. Because in a couple of years IoT and IIOT will cease to exist. Most people will instead talk in terms of connected products, experiences and intelligent machines/manufacturing systems
Security is a defensive measure. How does that transform manufacturing and make it better and more productive?
I am not sure why this is even listed.
Finally. This is a valid technology that has the potential to transform certain segments of manufacturing. Some of the most exciting work is being done with next generation materials (nano-materials) that are ideally suited for additive manufacturing.
Augmented Reality (AR)
AR is certainly an exciting and transformational technology for games and you know what. The authors also correctly note that AR can be transformational in product development, training and service. But before that happens it has to be cheap, good and easy to adopt. It’s not there yet.
Overall, the Industry 4.0 paper lacks intellectual rigor that is expected of a BCG paper. It has taken an unwarranted and excessive technology slant and failed to take into consideration other essential aspects, including new materials, new manufacturing processes, transportation trends, regional economic and human resource trends and geopolitics.
Zero lead time, zero inventory and zero waste are aspirations of Toyota Production System. Of course, they are physically impossible to achieve but those aspirations shape the questions asked, decisions made and how production activities are organized in the Toyota Production System. Toyota in fact managed to reduce lead time between a custom order and production to 5 days back in 1999. It is an amazing accomplishment even by today’s standards.
The idea of making and delivering personalized products to meet individual tastes has been an alluring one for product companies. Companies like Harley-Davidson, Dell and Motorola recently have offered some level of personalization for products to meet individual tastes. Latest example of Coke allowing names on personalized bottles also points to a future where customers can even personalize small and ephemeral purchases. Time will tell if it catches on and if this level of personalization will be a competitive advantage to companies that can support it.
Nevertheless, it is important that product companies be prepared to support some level of customization and personalization, especially in consumer markets. Companies such as HP and Apple use product design and manufacturing strategies such as postponement to support customization. Customization and personalization does not come easy and has a significant impact on supply chain and manufacturing strategy and performance. Supply chain and manufacturing infrastructure needs to be designed and implemented such that customization is possible and order-to-delivery lead times are reasonable. In addition, the story of Dell offers a cautionary tale for going over-zealous on personalization and configuration. Dell essentially lost marketshare as PCs turned into commodities, customers de-valued configurability and more variety was widely available in electronics stores with zero lead time. In consumer markets companies have to be especially vigilant of changing customer preferences and balance must be struck between lead times, availability and personalization.
Since most companies are not Toyota as far as Supply Chain Planning and Production Scheduling is concerned a combination of forecast based and order based production is most appropriate. Many companies in consumer products sector (food, electronics, etc.) have to necessarily use forecast based production planning and manufacturing execution systems because customer purchases are based on availability on the shelf.
On the other hand companies that manufacture and deliver customized products with short lead times can use order based production planning and manufacturing execution processes. This should be the case when order-to-delivery cycle time is greater than production cycle time. The Toyota Production System strategies that include single piece flow and mixed production are also ideally suited for these companies.