Steam engines sparked the first Industrial Revolution, electricity energized the second, and early automation and the assembly line powered the third. Now, the fourth (often called smart manufacturing) is being shaped by artificial intelligence, advanced analytics, the internet and real-time data.
Smart technologies are transforming manufacturing. It starts on the factory floor, optimizing processes through advanced instrumentation, automation, robotics and human-machine decision making.
The Clean Energy Smart Manufacturing Innovation Institute (CESMII) (pronounced sez-ME) was born about four decades ago, around the same time as the introduction of digital controllers and early AI systems.
CESMII is a US Department of Energy-funded public-private partnership and plays a pivotal role in promoting smart manufacturing and making it accessible to all manufacturers. CESMII’s focus areas include technology development, education, workforce development, and fostering a national ecosystem of leaders and innovation centers.
We recently talked with John Dyck, CEO at CESMII, about the challenges in manufacturing and the roles analytics and artificial intelligence will play in the future of smart manufacturing and the global competitiveness of US manufacturers. You can watch a short clip of Dyck below or read further for our full interview.
The case for smart manufacturing is incredibly compelling. Can you explain why the concept is so important?
John Dyck: US manufacturing has been characterized by innovation for centuries, but the innovation that’s taking place now is in pockets. And the innovation we talk about is unbridled innovation, one use case at a time, and the data silos and stovepipe architectures that were created because of that. Fundamentally, that’s what’s kept the US from adopting smart manufacturing capabilities to the same extent as both Europe and Asia.
We believe that a smart manufacturing mindset is key to improving the US’ competitiveness in manufacturing in the same way as the quality mindset and the quality movement from four decades ago. Back then, we became a nation obsessed with quality.
We became a nation obsessed with continuous improvement, with safety. The idea with smart manufacturing is to digitize our quality, productivity, maintenance activities, workflows and business processes.
For nearly all of US manufacturing’s known history, our worker productivity has steadily increased – until the last decade – which ironically coincides with the fourth Industrial Revolution and the expectation that there would be significant value creation. The reality, though, is that worker productivity began plateauing and then declined during the last decade. There was a realization, finally, post-pandemic that manufacturing and supply chains are dependent on data and that organizations that are further along on their journeys to smart manufacturing have been more resilient and productive.
What technologies are overhyped for manufacturing, and are there technologies that maybe deserve more hype?
Dyck: That’s a great question. There has been some anticipation in smart manufacturing around AI and machine learning for functions such as the connected worker and augmented reality. I can’t really call it overhyped though because it depends on where you are as a manufacturer. There are two ends of the spectrum, and CESMII needs to support both sides equally well.
The value and potential value realization of some of these new capabilities and technologies depend on some basic infrastructure readiness that broadly isn’t there. Well over 90% of US manufacturers are small and medium-sized manufacturing organizations, that have little access to these technologies. If you were to look at where they stand from a readiness perspective, they’re barely entering what we might refer to as the third Industrial Revolution.
On the other hand, the larger manufacturers have invested in the data infrastructure and information modeling and are ready for some of the significant new capabilities.
When I look at the notion of piloting these important new capabilities, I think many organizations struggle beyond the proof of concept. And that speaks to the fact that we have work to do on how to democratize these technologies. That’s an important word in our CESMII dictionary.
Democratization encompasses the idea that we have to reduce the cost and complexity by collaboratively solving problems and not just assuming that something we did once in a confined proof-of-concept will automatically scale. There are nuances around how we can capitalize on some of these “overhyped” capabilities. Both the analytical basics that are step one for most manufacturers and these incredibly powerful new tools are vital for our approach, for our investment and for the work we’re doing to democratize smart manufacturing.
For the past three years, there has been so much disruption, especially in the face of the pandemic. What are some things you’ve seen organizations do to be more resilient in the face of all this disruption? Would you say data and analytics have played a role in that resiliency?
Dyck: I saw some incredible heroics at the beginning of the pandemic as some manufacturers saw their demand skyrocket and some saw it fall away precipitously. So much of the supply chain disruption and the day-to-day tools manufacturers invested in on the supply chain side fell by the wayside when it came to the need to respond to something so extreme.
As manufacturers work on resiliency, it’s becoming clear that we need more systematic approaches for how we look at data. We need better analytics, and we need to be more collaborative in our approaches with suppliers. The reality is that the predominant form of communication with manufacturing suppliers during this disruption was still faxes, emails and phone calls. That’s not acceptable. Building infrastructure that allows suppliers and manufacturers to collaborate in real-time so we can apply more systematic analytics to the data is central to how we think about digital transformation, smart manufacturing and more productive US manufacturers.
I think we’re seeing the beginnings of the regionalization of supply chains. We’re seeing that as a response to what happened in the last three years; it’s the beginnings of “re-shoring.” But all of that requires a more productive manufacturing environment and not the reality that we’re less productive today than we’ve ever been. These are vital areas that we have to focus on as a nation and solve together.
How have you seen analytics help organizations improve their productivity?
Dyck: When we move toward the digitization of data, real-time data on the shop floor, there’s been an almost universal and significant uptick in productivity. When the systems are first turned on, people don’t believe the data that’s been collected. Lean manufacturing systems are often paper-based. They’ve indicated a level of productivity, but when you get the digitized data, the actual productivity is often 20% to 25% lower than what they assumed when they were looking at the data manually.
And so the idea of digitization – removing human bias, removing the basic mistakes that happen when you’re aggregating and collecting data from multiple data sources on the shop floor in a digital way – represents a huge productivity curve. That simple act of digitization reveals productivity opportunities to the tune of about 25%.
Where do we go from here?
Dyck: The next step in smart manufacturing is the digitization of quality, productivity, maintenance activities, workflows and business processes by creating a mindset of smart manufacturing. In that culture, we demand data. We demand analytics. We demand objective realities, not subjective, and data-based instead of gut-feel decision making.
We must help all manufacturers create an ecosystem that includes technology providers, machine builders, implementers and systems integrators. We must help them move toward the smart manufacturing mindset.