Is My Smart Factory Smarter Than Yours? It’s Hard to Say

The concept of the “smart factory” is hot in the manufacturing sector, but the precise meaning of the term and the prospect of how to build one can be elusive.

What those terms mean is something of an open question. Each has become more of a sort of marketing slogan with evolving meanings than a definable signpost for high-tech manufacturing. And terms referencing intelligence or, for that matter, industrial revolutions, tend to become ever-more grandiose or more generic over time. 

Such is the case with the second of the terms in this list, which came to life in 2011 as a German-government-backed Industrie 4.0 initiative to help the nation’s manufacturing sector maintain its competitive edge. The phrase itself hints at a cyberphysical, systems–driven industrial revolution, but, with its “.0” suffix, points to the level of change one would find in an incremental, but substantial software upgrade. It also conjures to mind the Web 2.0 that largely receded from memory since it popped up around the turn of the century and peaked around 2007. And at this rate, Web 3.0 has failed to become a mainstream buzzword. 

Deloitte pointed out in 2017 that the term “smart factory” indicates more of a journey than a destination. “[I]t represents an ongoing evolution, a continuous journey toward building and maintaining a flexible learning system—rather than the “one and done” factory modernization approach of the past,” reads part of the article “The smart factory: Responsive, adaptive, connected manufacturing.” 

Current Factories of the Future

There are, of course, a growing number of reports of factories and industrial facilities that seem to indicate where the manufacturing field could be headed. 

For instance, in 2017, Bloomberg reported on a steel mill in Austria that churns out 500,000 pounds of steel annually with just 14 workers. 

The facility is by no means unique in terms of its prodigious use of automation. The Japanese robotics company FANUC has operated a so-called “lights-out” facility since 2001, according to a 2003 Fortune article. The article reported that robots churned out some 50 robots per 24-hour shift. “Not only is it lights-out,” said Gary Zywiol, then a Fanuc Vice President. “We turn off the air conditioning and heat too.” 

The FANUC facility has a certain similarity to Philip K. Dick’s 1955 short story “Autofac,” in which machines self-replicate. In that dystopian work of fiction, however, the automated factories operate against the wishes of a limited number of human survivors. 

As much of the manufacturing industry across the world grapples with rising labor costs and labor shortages, the prospect of completely automated facilities continues to be en vogue.  

Some companies are moving in the opposite direction, however. Toyota has replaced robots with human workers wherever possible. “The automation process will progress, generally speaking,” Mitsuru Kawai, Toyota’s Head of Manufacturing and Executive Vice President told Roland Berger. “But when we use robots, they’ll be trained by people who know what they’re doing.” 

Tesla, conversely, planned on completely automating its so-called “Alien dreadnought” factory in Fremont, Calif., but its leader Elon Musk had a change of heart after realizing humans were “underrated” as he quipped in 2018. While Tesla’s factory still makes extensive use of automation, it also highlights how such facilities can drive demand for specialized white-collar positions such as battery algorithms engineers, computer vision scientists, and deep learning and machine learning engineers. 

A Happy Medium — at Least in Concept

As the Toyota and Tesla examples illustrate, a number of facilities are deploying automation in conjunction with automation, gradually optimizing their operations in the process.  

Christian Lutz, Chief Executive Officer of the IoT data management firm Crate.io, has seen how the march toward autonomous self-optimizing production lines can transform industrial companies. One of its customers, the plastics molding firm ALPLA, has rolled out automation technology with machine learning capabilities in several of its North American facilities. “It’s a dramatic change,” Lutz said. “For example, before the system was deployed, a shop floor worker would walk about 10 kilometers a day between the production lines with a clipboard and tick off boxes.” Roughly speaking, a worker would walk to a machine, verify it was working correctly, check a box and move onto the next machine until a problem became apparent. To date, the company has rolled out the system in 11 factories in North America. 

And if there is a glitch, it can affect potentially thousands of products. “That means you need a real-time reaction system, but also a system that hopefully predicts the problem,” Lutz said. “Maybe you see a trend building up. You see the tolerance, and you get alerted when the system goes out of tolerance. Hopefully, the system would see this trend coming and it would alert you one minute before.” 

While many manufacturers continue to make greater use of automation and tools such as machine learning, the overall trend has a long history. Mid-sized companies in German-speaking Europe known as Mittelstand have embraced automation for decades to remain competitive in a world where cheap labor is always available, somewhere. “In Europe, you have these mega-optimized factories that are already super automated,” Lutz said. “They’re all sensored up. They are collecting information from the shop floor. But the information is only used to visualize the current state of production.” And if a problem pops up, human workers must scramble to fix it. 

Other manufacturers across the world have embraced a similar methodology, but the smart factory trend changes the calculus. It promises to enable manufacturers increasing capabilities to predict problems before they cause downtime. Granted, the technique is often more-difficult to deploy than anticipated. It tends to work the best for organizations with considerable data on a significant number of machines, said Hala Zeine, SAP’s President of Digital Supply Chain and Manufacturing. 

Another hurdle for predictive maintenance is the question of where the data processing should be located. What should be processed directly on the shop floor — at the so-called “edge”? And what should be sent to the cloud? “To get this balance right is the Holy Grail in smart factory,” Lutz said. “In the long-run, pushing everything to the cloud doesn’t work from a cost point of view.” 

There are a number of companies pushing analytics to the edge and aggregating data to the cloud. “But what people often forget is: once you aggregate and compress the data, for example, to ‘max,’ ‘min,’ ‘outliers,’ ‘average,’ and stuff like that, you lose the ability to run data science,” Lutz said. “You need the raw data, but where do you store that?” 

A Difficulty with Definitions

With its emphasis on cyberphysical systems, the term Industry 4.0 is an arguably even more ambiguous term than smart factory, but its clearest reference is its hint at a forthcoming industrial revolution. Whether the term “Industry 4.0” has staying power a decade from now is uncertain, it is difficult to gauge the current state of the broader industrial sector is a bit difficult to decipher. 

But if there is an industrial revolution underfoot, it may be different than you imagine. Depending on whom you ask and where you are looking in the world, manufacturing productivity is either sluggish or has been trending incrementally higher in recent years. Germany may be a trailblazer when it comes to smart manufacturing, but signals are mixed regarding the speed of transformation for the nation’s industrial sector. Its manufacturing industry, along with its powerhouse automotive sector, is “in trouble,” as The Guardian recently put it, citing its vulnerability to global economic uncertainty and trade tensions. On the other hand, the Dieselgate scandal has led its automotive industry to double down on innovation, beefing up research and development on electric and hybrid vehicles. Volkswagen also plans on spending some €10 billion on a high-tech lithium-ion battery manufacturing facility and to became the first auto company to deploy 3D printing at scale.

German carmakers, like their international competition, have also invested in ride-sharing, autonomous vehicles and multimodal. Daimler and BMW, for instance, are jointly investing €1 billion in a ride-sharing venture.

VW’s changing business priorities is a microcosm of many industrial firm’s quest to reinvent themselves. 

And for those solely looking at, say, the productive sector of the manufacturing sector as evidence for industrial transformation are not seeing the whole picture, said Martin Davis, Managing Partner at DUNELM Associates Ltd. “It reminds me of the quote attributed to Henry Ford,” Davis said. “If I had asked people what they wanted, they would have said a faster horse.” 

Erik Josefsson, Head of Advanced Industries of Ericsson, has a paradoxical-sounding view of the state of the manufacturing industry. “You actually don’t know the change until you have actually had the change,” Josefsson said. “Because when you’re in it, you don’t really realize it in the same way.” 

Part of the challenge is the expanding definition of terms. Any technological term with “smart” preceding it likely lacks a firm definition. And then the Industrie 4.0 framework is also changing. “You could say in the first place, it was more about getting connected. We started to talk about ‘cyber-physical systems.’ And then the whole concept of digital twin emerged,” Josefsson said. “But in reality, now we’re taking digital twins to the next level. So the definition of digital twin has changed.” The same thing applies to so-called factories of the future, which, in a literal sense, will always remain out of reach. Josefsson added: “We are increasing our expectations and asking: ‘When is that revolution going to happen?’”  

Written by:  Written by Brian Buntz, Journalist, for IoT World Today.