Factories of the Future: Five Expert Perspectives

How will Factories of the Future be organized and managed? What are the implications for the way data will be used, production networks operated, and manufacturing enterprises structured over the next decade? Five leading academic members of the Manufacturing Leadership Board explore the possibilities

The following extracts from our conversations reflect some of their most significant perspectives.

Included in this article are leading academic insights from Dr. Jim Davis, Vice Chair, Clean Energy Smart Manufacturing Innovation Institute (CESMII) and Vice Provost, Information Technology and CTO, UCLA; Dr. Larry Lapide, Research Affiliate and former Director of Demand Management, MIT Center for Transportation & Logistics; Dr. Paul Christodoulou, Principal Industrial Fellow at Cambridge University’s Institute for Manufacturing (IfM), U.K.; Dr. Jay Lee, Ohio Eminent Scholar, L.W. Scott Alter Chair Professor, and Distinguished University Professor at the University of Cincinnati; Dr. Detlef Zühlke, Executive Chairman of the SmartFactoryKL Technology Initiative, Germany.

Next-Generation Agility

Jim Davis:
One of the key factors driving the transformation to factories of the future is the changing nature of demand. The customer today wants to have much more variable value-add, much more specialty, or much greater precision or managed precision, around the different attributes of a particular product. There is strong economic impetus toward precision products that are purpose-built, safe, and more safely produced, environmentally-neutral, and manufactured with far less energy and material throughout the value chain. But they also want to get these individualized products faster than ever before. For manufacturing, the objective is to be much more directly responsive to what the customer wants. But the business of specialized features starts taking on more and more of a character that looks like a low-volume mix; it’s just that you’re doing many, many more low-volume mixes. We already see this in healthcare and pharmaceuticals with precision medicines and therapies where they’re focusing on smaller units and smaller reactors and plug-and-play operations to get more precision stuff. Manufacturing is focused on productivity and performance now, but we’re starting to see this kind of precision product and facilities modularization happen even with larger organizations and suppliers. I see a strong trend toward this modularization of processes and materials, followed by the modularization of manufacture of the product itself. That’s not the kind of agility your father talked about. It’s “just in time,” but on steroids. It’s the next generation of agility.

Dr. Detlef Zühlke, Executive Chairman of the SmartFactoryKL Technology Initiative in Germany, former Scientific Director of Innovative Factory Systems at the German Research Center for Artificial Intelligence (DFKI), and Chair of Production Automation at the Technical University of Kaiserslautern. He is a guest professor at the University of Cincinnati and was awarded an honorary doctorate from Sibiu University, Romania. He was also awarded the Rhineland-Palatinate Federal State’s Order of Merit last year by German Prime Minister Malu Dreyer for his work introducing digital smart automation into factories. He is also a Board member of the Manufacturing Leadership Council.

The Learning Factory

Jay Lee:
If you look at factory design today, it’s based on product launches. Look at airplanes and models like the 737, 747, or 757. It’s the products that are driving the factory set up. Then you have a lean system that is developed for that function. Then you have the IOT system to support that function. Then you have people training to support that function. It’s a product-driven factory today. But tomorrow, you’ll see more products driven by customers, so it will be a more customerdriven factory. First, the customer will engage with you directly in design and perhaps production. Second, the factory is no longer making just a product alone. It has to start making knowledge, because we now live in a digitized, well-connected world. Machines, using data, can provide evidence. The machines in the factories of tomorrow will have more “evidence” about how a product is made, rather than just today’s “experience” of how it’s done. People retire and we have a new generation coming through. So we need to engage more and we need a factory to generate more evidence, more learning knowledge. Tomorrow’s factory will become a learning factory, not just a product-making factory. As people become smarter, and machines become smarter, people will be able to learn from machines. Third, we need to incorporate more advanced devices in the factory design. Young people are coming to factories now and they need excitement. They need to feel they are involved in something they believe can be part of their life. Machining used to be very boring. Now machines have what we call “cyber-physical” systems. They can show people how the part is made, how they improve it, how to make it better, how to waste less energy. This evidence-based approach is going to make people feel more desire to spend time in a factory. And with evidence-based factory intelligence, I think, learning can become a very interesting asset for a manufacturing company in the future, not just machine assets and factory assets.

Treating Data as an Asset

Jim Davis:

The whole notion of making data a key asset in the manufacturing process changes things dramatically. What does it actually mean to make data a key asset and then work this emphasis through the culture of the entire company? How do you manage it, collect it, and use it? It requires a change in the whole focus of facilities and how you deploy resources and effort. That is a significant mindset change for manufacturing not to drive everything from strictly a facilities focus, but instead, to think about facilities as a means to an end, where the data it creates is actually driving what you’re doing, how you’re doing it, and when you’re doing it. It is this mindset that opens the potential for managing material and energy, upstream and downstream effects, physical assets and cyber assets, and economic and market potential throughout the value chain.

Creating new organizational and architectural infrastructures are a key aspects of this data-driven trend. The infrastructure with which data is managed, both within companies and at a broader industry level, needs to change dramatically to provide the ability to bring data together effectively, so it can be analysed and deliver value in different ways. Companies are increasingly seeing value in interconnecting, ingesting, provisioning, and orchestrating data for manufacturing action up and down the value chain. You just can’t use data in a comprehensive manner and not have the organization organized around it. If companies are increasingly going to create trusted business relations involving data (virtual businesses) across operations, or line operations, and then eventually supply chains and across companies, infrastructure, security and trusted data management will play a huge role. Perhaps there is huge advantage, from a business standpoint, of having a place where that data can be brokered and shared in a trusted way to help form these virtual collaborative businesses of the future.

Multi-Dimension, Integrated, IOTT

Jay Lee:
The way I look at it, IOT is basic stuff right now. I think IOT needs to become IOTT, the “Internet of True Things”. True things means you truly understand why you connect. Today, we’re
just trying to connect things and we don’t know why sometimes. It doesn’t make any sense. Machines should connect not just because they can be connected, but because they connect for value. Then they can focus on creating more effective data, more effective learning, more effective usage, and more effective cost. The next level for the smart factory is multi-dimension integrated IOT. IOT should not just connect simple data or measure only one parameter, it should be a multi-dimension connecting sensor. For example, take machine vibration. When I measure it, a multichannel Edge Device can intertwine vibration, acoustics, and current together. So it’s sent out as a correlated feature, rather than individual signals. With 5G environments, I will have the volumes to intertwine multi-signal, multi-dimension information. But to intertwine them, we need to better engineer multi-dimension IOT so that’s it’s no longer data sensing, or parameter sensing; it’s information sensing. Today, I take vibration data, I take temperature, I take acoustic signals, and then I go off into another system. Tomorrow, I will be able to sense all these together in a meaningful way so I know exactly if the machine is not stable and why. I won’t have to do all the analysis anymore. Integrated multi-dimension IOT will give us much more cohesive visibility, learning, predictability, and prescriptive decision-making in tomorrow’s factories.

Virtual Manufacturing Models

Jim Davis:
The factory of the future is becoming more of a virtual business agreement, as opposed to a facilities-oriented-driven factory. The notion is of a virtual agreement that, then, manifests itself in a set of physical operations that are driven by that virtual agreement. A good example of this is a process in one of the world’s leading materials companies. They in-house half a project and then they contract out the other half. It is actually easier for them to contract it than to build it on the line themselves as they can draw on multiple areas of expertise around a particular material and get the whole line operation setup faster. It’s a virtual line, but to all intents and purposes, it looks like a coherent factory.

Dr. Larry Lapide, Research Affiliate and former Director of Demand Management at the MIT Center for Transportation & Logistics where he also managed the launch of MIT’s Supply Chain 2020 Project exploring the future of supply chain management. His is a Lecturer at the University of Massachusetts, and a Board member of the Manufacturing Leadership Council.

Digital Supply Chain Visibility

Paul Christodoulou:
We take a factory network view, as well as a factory view, when we think about the future of manufacturing, and look at how new digital technologies may impact the full end-to-end supply network. First, you’ve got a set of issues around integrating the inbound supply chain. That’s about making sure all your suppliers and sub-suppliers are all linked up and integrated, so you’ve got better visibility from all the data that’s now available. It’s no good optimizing what you do in your own factory if you’ve got a shortage of one essential part because one of your suppliers didn’t work out. Second is the outbound view, and the integration of all the aspects here too, which in some ways, is more challenging and complex. You’ve got to create links all the way through your direct customers and channels, all the way to the end customer and consumer. And you’ve got to consider things like last-mile logistics and some sort of feedback channel from the customer, whether it’s through a web-based product configuration approach or a customer interface. There’s still a lot more that can be done here, I believe. The final part of this is the total end-toend view. When you come back to two of the key aspects of digitization giving us value – integration and visibility – a lot of the value must be associated with getting a better sense of this total end-to-end view. That could be from a strategic design perspective; how you use digital tools to inspire change in the whole design of the network over a ten-year period. It could be on a disruption-checking basis; getting a clear end-to-end view of possible disruptions and being proactive in preventing them. It could be from a quality perspective; feeding back quality issues right back through the chain to the factory and making sure the feedback loop is properly in place. Or it could be around the linkage with product lifecycles. That’s perhaps the most complex thing. So I think the impact of digital transformation will be much more than just within the factory itself, but in the creation of factory networks and digital supply chains that are closely aligned and responsive to next generation product lifecycles that will drive end-to-end value.

Project-Based Enterprises

Larry Lapide:

What makes a manufacturing company? Basically it’s a group of entities that have expertise, that are getting together for the purpose of a mission, which is to make and deliver a physical product. It may have existed for a hundred years or more, but it’s really a project to produce a certain thing, whatever it’s successful in, or to try to develop and make new things. But the model I think will emerge in the future for manufacturing is more like making a movie. There’s a producer and he brings on people who don’t work for the company; they work for a project called “the movie.” Or look at the Uber guys. They’re all contractors too. So I think we’re getting closer to everybody becoming contractors and where the business becomes much more project-based and connected through real-time technology. If I decide I’m going to make a product, I get a group of contractors together. One of those may be the guy or woman who runs the plant – or a group of people running multiple plants. Now we’ve got a design issue. So we need people to design the software, design the hardware, and they’re all contract-based as well. So I’m putting together this thing for a product, it needs software, it needs hardware, we need to design everything, we need the right modules. So we’re really bringing a supply chain together for this product. We may only do that for five years if the product doesn’t work out. We’ll disband after that. But if the product works out, hey, we’ll be together and doing this for a long time. I think this is closer to the project-based model of the future for manufacturing. It’s an on-demand approach. The whole concept of a manufacturing company will then be different and the organizational chart won’t look the same anymore.

Composite IP Challenges

Jim Davis:
One of the biggest challenges I see for the future is how we approach intellectual property. The intellectual property that we’re talking about with the new kinds of integrated systems is what I call “composite IP”. It’s the notion of starting with a sensor technology, and then someone else bringing in a modeling technology, and someone else bringing in an optimization model etc., to form an overall system. Consider an acoustic sensor combined with a data-driven, machine-learned model to build a way to listen for the quality of a part during a machining step. There is great value in this composite system, which can become a product in itself and used for another similar application. There’s a lot of new development and new IP in each aspect, of course, but there’s a lot more value in the resulting integrated IP. Unless we can develop a business model in which composite IP delivers value to all, the rapid deployment of these new IT and data technologies could become a real problem.

Adopting a Digital Attitude

Paul Christodoulou:
One of the key enablers for the future of manufacturing is what we call a “digital attitude.” Naturally, many companies are currently focused on new technologies and systems to improve productivity and flexibility in their factories, which, of course, is important. But, in order to make a success of this, one of the leading capabilities that’s needed now is to find people, either through training or the right recruitment, who have a good understanding of the digital environment but in a very holistic sense. Not just a good understanding of the systems and the software capabilities, but how those link to business systems, business processes, machinery, products, and being able to visualize the underlying value proposition of this digital shift. That’s very different from technical experts or technology experts, and it’s very different from your traditional general business manager. They are a new kind of digital champion who can drive a corporate-inspired shift to digital transformation. At the end of the day, the advantages of the digital shift are around things like integration, visibility, automation, of course. But in order to understand the possibilities for the future of the business, there needs to be a very high-level view of those issues and how they link with the technology enablers. The real challenge is then to link them to business model innovation and creating value. That’s the gap that everybody’s facing.

Enlightened Leadership

Detlef Zühlke:
In the past, manufacturing companies were mostly made up of small, very hierarchical, pyramid style structures. But in an increasingly connected world, I think manufacturing companies will be more like an open network of systems. Many organisations will have to dissolve their hard-wired, pyramid structures and create flatter approaches. That means preparing people to be able to think in terms of the whole business environment, or the whole factory, and not so much just their own little area that they’re responsible for. The key to achieving this, for me, is recognition by management that significant changes will come and how they should react to all this – at a technical level, at an organisational level, and at a business model level. New digital technologies will drive a lot of this change, but the future of manufacturing is not just a technical question at all. It really needs to be driven by management to push these things forward in their own organizations.

Embracing Disruption

Jim Davis:
Transformational change has to be disruptive, otherwise it’ll take forever if you just try to tackle it one piece at a time. It’s important to tackle the training and education, the investment, the IP, and the technology in a comprehensive and coordinated way. It is better to take small steps in all areas together, than to take a large step in one area and then sequentially try to take on another. If you try to do the technology without the others, you just won’t get there.

This article is based on interviews conducted by the Manufacturing Leadership Council’s executive editorial team – David R. Brousell, Jeff Moad, and Paul Tate.