Episode Show Notes
Episode 4 features Dr. Jim Davis, Vice Provost of IT at the Office of Advanced Research Computing at UCLA. Jim defines smart manufacturing, explains how it’s evolved over the past two decades, and provides some unique examples of how smart manufacturing is being used today.
Jim Davis is Vice Provost of IT at the Office of Advanced Research Computing (OARC) with broad responsibilities for data and technology solutions in support of UCLA’s research mission and its broadly defined communities engaged in campus, local, state, national, and global impacts of digital research and scholarship. Jim co-founded the Smart Manufacturing Leadership Coalition (SMLC) and spearheaded UCLA’s leadership role in forming Department of Energy’s (DOE) Clean Energy Smart Manufacturing Innovation Institute (CESMII), the 9th Manufacturing USA Institute, and the 3rd DOE institute to be awarded. CESMII’s program and administrative home is with UCLA and through Jim’s Office of Advanced Research Computing. Jim led the Institute’s first business and technology roadmap and is currently involved with CESMII’s Smart Manufacturing Innovation Platform infrastructure and its use for data and modeling applications in manufacturing operations and supply chains.
Highlights
00:00:00 - Introductions
00:01:25 - Definition of Smart Manufacturing
00:06:36 - The technology’s breadth of applicability
00:07:12 - Examples of improvements from adopting technology
00:15:11 - Discussion about ROI
00:22:29 - Smart manufacturing tools that bridge the gap between IT and OT
00:23:42 - Adoption and barriers
00:29:20 - How Smart Manufacturing has developed over the past 10 years
00:31:29 - Security considerations
00:33:41 - Projections for wider use and ease of implementation
00:37:37 - Key cultural elements and technical capabilities needed for ROI
00:40:20 - Next evolution of technology and trends
00:43:36 - Company preparation
00:46:50 - Summary
Transcript
Gregg Profozich [00:00:02] In the world of manufacturing, change is the only constant. How are small and medium-sized manufacturers, SMMs, to keep up with new technologies, regulations, and other important shifts, let alone leverage them to become leaders in their industries? Shifting Gears, a podcast from CMTC, highlights leaders from the modern world of manufacturing from SMMs to consultants to industry experts. Each quarter we go deep into topics pertinent to both operating a manufacturing firm and the industry as a whole. Join us to hear about manufacturing sectors' latest trends, groundbreaking technologies, and expert insights to help SMMs in California set themselves apart in this exciting modern world of innovation and change. I'm Gregg Profozich, Director of Advanced Manufacturing Technologies at CMTC. I'd like to welcome you. In this episode, I'm joined by Dr. Jim Davis, Vice Provost of Information Technology at UCLA Advanced Research Computing. Jim and I discuss Smart Manufacturing, its definition, how it's evolved over the past two decades, and some unique examples of how Smart Manufacturing is being used today. Jim closes with his insights on how organizations can get the most value out of Smart and what he sees is the future of Smart Manufacturing. Welcome, Jim. It's great to have you here today.
Jim Davis [00:01:17] Gregg, it's good to connect up. We certainly have been talking about this a long time, haven't we? Was trying to count the years. It's probably five or six years at this point.
Gregg Profozich [00:01:25] It probably is. Early 2015, I think, was the first time we started having Smart conversations. I know you've been at it a lot longer. I'm really excited about our conversation today, though. If you think about it, the manufacturing world has been experiencing yet another shift in recent years, often referred to as the Fourth Industrial Revolution, the other three revolutions being number 1.0, the shift to mechanization and steam power from animal-driven power; the second one, the shift to mass production and assembly line; the third one being the integration of computers into operations technology — for example, PLC logic — and number four, the emergence of cyber-physical systems and IoT. Smart Manufacturing falls under this Industry 4.0 umbrella. But the term smart is used in many, many different ways these days, from smartphones to smart refrigerators to smart cities. For context, let's talk a little bit about what you define as Smart Manufacturing. Share with our listeners your view on what it is and what it isn't.
Jim Davis [00:02:16] Well, there's a couple things to get to the definition. First of all, I have to laugh about it a little bit. Back in 2005 when we first coined the term Smart Manufacturing, we googled it, and we got less than 10 hits. Today, if you google it, there's millions and millions of hits. So. there was a time where we had hold of the definition, and since that time it's expanded into many, many different versions and variations of definition. To get at the definition, I thought a little bit of history might be useful just to get at where the definition came from. Then that will give some context going forward to the definitions we're using today. There's a couple of connection points. One is, from my perspective, the term Smart Manufacturing was literally coined in 2006 at a National Science Foundation workshop, and at that time, it was literally called Smart Process Manufacturing. As we expanded and the role expanded, we shortened it to Smart Manufacturing. There was an interesting parallel in 2005. Germany started a similar initiative completely independently called Smart Factory, and a couple years after that, they renamed it Industry 4.0. The efforts with Germany and Industry 4.0 have been parallel. So, Smart Manufacturing and Industry 4.0 parallel themselves in many, many respects going forward. If you go back to 2005, the reason Smart Manufacturing got considered was that was the time when the word cyberinfrastructure entered the vocabulary. This was the time when National Science Foundation and the research community started looking at what's the role of cyberinfrastructure in practical applications. At the particular time, cyberinfrastructure meant what can I do with data modeling computation when I don't have to worry about its location or having the facilities, and what can I do with the fact that there's now a network, and I can exchange information, I can aggregate resources, look at other resources, and so forth? One other really important point, I think, to keep in mind is that the Web really didn't start till 1995. So, in 2005 when Smart Manufacturing, or Smart Factory, or Industry 4.0 started, the Web was only 10 years old. If you fast forward to today 15 years later, we're really at very early stages of really taking these kinds of technologies on. They've moved very, very quickly. But to what you were saying, it's exciting because it's moving fast; it's accelerating; and we're finally, from my perspective, at a point of readiness that these things are having real, true, practical value. So, I still haven't defined it. So, our elevator definition goes something like this. We just simply say, "How do we get the right information at the right time in the right form in the right place with the right technology to the right people so we can make decisions or take action?" As simple as that. So, it's a data-centric, data-focused activity with respect to manufacturing. The immediate follow-on question we generally run into is, "Well, to do what?" Most applications we're seeing right at the moment have to do with modeling and analyzing, but they're very quickly moving into monitoring, very quickly moving into controlling and diagnostics. Where we really want to get to is prediction, optimization, and what I call self-interrogation, but this is under different headings, such as health monitoring, machine monitoring, machines telling us when they need to have maintenance, all sorts of things like that. Then the last ingredient in our definition is that it spans the whole space. So, from a manufacturing standpoint, it spans from sensor and human-centered information all the way to the supply chain. So, how do we put all three of those components together? So, I hope that gives a little bit of a definition. But the historical reasoning and how it got here, I think, is valuable, as well. At least I hope it is.
Gregg Profozich [00:06:20] Oh, I think so. I think definitely. I think that the key takeaway if I'm listening well today, is that it's the litany of rights. I want to put the right data and the right information in the hands of the right person so they can make the right decision right now.
Jim Davis [00:06:35] Exactly.
Gregg Profozich [00:06:36] So, Smart applies to just manufacturing, it applies to any industry? What's the breadth of the applicability of this technology?
Jim Davis [00:06:44] When I was describing 2005 and when we were first using the term Smart Manufacturing, this was before other things were smart. So, if you look forward and look at today, as you said before, there's smart cities. That came in afterwards. There is smart health. We really are talking about the application of data and broader and very scaled ways to create all kinds of effects that improve, in our particular case, manufacturing.
Gregg Profozich [00:07:12] Okay. So, you touched on it a moment ago, I think, but maybe we want to go into just a little more depth and probe a little deeper on the types of issues where Smart can be applied and can share some real-world examples, maybe, of some things that have been done in the Smart Manufacturing space that had some tangible results.
Jim Davis [00:07:28] Yeah, I was hoping to get to some examples. That's truly the best way to describe this, although it's useful to come out and talk about the broader viewpoints. One of my favorite examples is with a small manufacturer that did nothing but make specialty corrugated packaging materials to order. So, there was many different kinds of products, but it all used the same or similar materials. The raw material that came into this small operation was basically these big rolls of heavy paper, a cardboard-like paper on a roll. They unroll it, and then they put it in different steps and different machines going forward. One of the things that kept happening on this process is that they had to shut it down while they changed out the roll, and they just watched when the roll ran out. They weren't watching the volumes of these products. They had to shut down to change a roll out. Then the question became: could they do a little better with that? So, in this particular case, what they did was put a single weight sensor on the roll mechanism. What they started out with was just simply what's the weight of the roll and how is it changing over time so they could take a look at speed of process. That alone helped them understand when that roll was about to run out. Then they planned it a little more carefully, and they started changing that roll out at better times relative to the production. When they got used to that and used to that data, they started expanding it, and they started saying, "Well, why don't I tie it to my different products? Why don't I start understanding how much of the material is actually used on a product basis? What if I could schedule these rolls and their amounts on these rolls, their two different volumes, two different orders, and so forth?" So, now they tied that weight sensor and that speed of process, and information into their ordering system. They got much, much more effective operations moving forward, started seeing places where they could optimize the process. This, literally for them, turned out to be about a 20 percent economic value just with one single sensor. The reason I like this example is we tend to see for small, medium, and large companies just getting started with one sensor and beginning to get the practice going with that data and getting that data set up is a very, very useful way to start. Then the complexity, you have to watch out for it. But there's just a world opened up with the different ways that you can use the data going forward.
Gregg Profozich [00:09:53] I think I heard you talking really about a crawl, walk, run approach. Don't try to do the big bang and sensor everything. Let's take a look at one critical thing. Let's start measuring; let's start getting comfortable with using data and then increment it out in ways that make sense. As you were talking, I had visions of balancing production to actual demand on an order-by-order basis. It's a takt time concept from Lean. We want to balance it out and not make additional inventory. So, I think a lot of these things are going to resonate with our listeners as we tie in that way. Maybe another example or two, and then I do want to talk about the economics of Smart Manufacturing and what industry can expect in terms of improvement potential and ROI. But let's talk first about a couple more examples maybe.
Jim Davis [00:10:37] A couple more examples are useful. This is a small company that makes a specialty material for the manufacture of glass. That material is shipped to an OEM that makes specialty glass. In this particular case, it was glass for the front end of computer screens. The large OEM and the small manufacturer discovered if they could get some sharing or exchange of some key properties on the material while it was still being made in the small factory, the large OEM could shift and adjust their processes to make sure that they could handle variations in the product quality, and the small company could make adjustments in their own process if they were seeing product spec going out relative to what the large OEM needed. So, it was much more real-time interoperability around the properties of the material for making that glass. But this is an example of a small company interacting with a larger company in a supply chain. If I could build on that example very quickly, this business of line operations and supply chains acting like line operations is a particularly important area in Smart Manufacturing. One example that we have that we worked very closely with was in metal fabrications. In this particular case, the front end of the process was a forging press that took steel billets and pressed them into a circular shape with a hole through the middle, because they made round artifacts with the center pressed out, like piping kinds of things. There was a heat treatment process in which that material went into its hardening process. That was a highly energy-intensive process. Then it went to a machine bay, where different parts were machined to high degrees of precision, and they made a number of different products and parts. From a Smart Manufacturing standpoint, what was done was on the forging press, it was sensored to begin to see when the center hole went out of spec when the diameter and the specifications started shifting so that it could be taken down and maintenance performed rather than letting it drift too far. On the heat treatment furnace, different products are going in. What they were doing was overtreating them. It was left heating much longer at much higher temperatures than what was needed. They were able to, with some better sensing, change the heat treatment process and save quite a bit of energy. On the machining side, by dealing with the specifications on the forging press, they didn't need to rely so heavily on the machining, which was forcing much more maintenance, much more pressing of the machines, and to higher volume, deeper, faster kinds of machining cuts. It saved on the equipment. It saved on the electricity and utilities on the machine side. If you put all three of those together, it was about a 20 percent savings economic value for this company, but it was across different aspects of things, machine maintenance to energy savings. But it's a good example of integration where the process beforehand was vertical operations that were not interacting very closely at all from a data standpoint.
Gregg Profozich [00:13:57] So, I think those are three great examples. In the first one, we had a company using data to improve their product-to-order connection and making a tighter connection there that allowed for some efficiencies and some savings. In the second example, we saw the integration of supply chain trading partners where two companies, separate entities, could begin working as a single line, if you will, and synchronizing their production based on quality levels and outputs. Then the third one was a maintenance and, really, quality example, where we talked about the machines being maintained correctly, and tying sensoring information into the quality of the product coming out and making sure that the parameters were set correctly for whatever was running at the time.
Jim Davis [00:14:36] Exactly. It's illustrative of sensors being used to improve the performance of an individual unit. It's also indicative of sensors being used to increase the precision of the parts on the machine side, and it's indicative of working across a line operation that would increase the productivity. From Smart Manufacturing we have put a label on all three of these. We call it productivity, precision, and performance. I use those terms over and over again.
Gregg Profozich [00:15:03] Productivity, precision, and performance.
Jim Davis [00:15:06] Those three show up all the time and in combination with each other.
Gregg Profozich [00:15:11] Okay. So, can we go a little bit more in-depth into the ROI piece? We know what Smart Manufacturing is now. We know where it's been applied. We've heard some real-world use cases. Talk about ROIs and potentials there.
Jim Davis [00:15:22] So, to talk about the ROI, you have to consider two different areas. One is the business economics and why are you doing this in the first place? Why does the company want to do this? I hope those examples illustrated some of the reasons. But it's also quite interesting to look to the future, and even the near future, because what we're seeing across all of manufacturing is the demands are increasing for higher precision, they are increasing for smaller lots that are more focused, more customized for particular customers. There is a greater need, quite honestly, in responsiveness — faster, cheaper, safer. The global side of this is very, very important. There's a lot of interest in market share not only in the US but globally. The environment plays a big role. How do you minimize the effects on the environment? If you can reduce energy, it's a real savings. If you can reduce material consumption, it's a real savings, and it helps the environment. Then, how do you do precision validation? I'm talking about both parts and materials, and parts and materials together. If you look at how manufacturing is trending, there is more demands on aspects to these where the data plays bigger and bigger roles. If you play these out, we're starting to see significantly across all industry segments, for all practical purposes, smaller upstream effects are having larger downstream effects. Things like tracking and traceability are becoming more and more important. Aspects like that interoperability example on the glassmaking. So, the supply chain effects are really important. Interoperability becomes very important in getting capability to deal with the data in the hands of the small, medium, and large manufacturers altogether is becoming more and more important. So, the economics of that are playing a very, very big role going forward. There's also the economics of the data side. On those examples which I just described there's a very, very big issue, a very big challenge. Most of the companies, if they're trying to do this at all, are building their own infrastructure; they're building out their own software; they're dealing with the vendors by themselves. If you add up the cost of this, about 70 percent of the cost goes to the IT infrastructure, and then another 25 percent goes to just working with the data to get it in the right form so you can understand it and interpret it for whatever you're trying to use it for. So, if you add both of those up, about 90 percent of the cost is just getting ready to use the data. So, we're seeing over and over again that because of this independence, company by company approach to this, for those who are even trying at this independence, is creating these silos that are just costing too much to do this. So, there's a real need to take lessons learned from the IT industry and apply those to manufacturing, to get these costs on the IT, the software, the management of data down significantly. So, there's a huge potential on the business side, but we have to get the cost down on the implementation side.
Gregg Profozich [00:18:29] Okay. So, if I'm a small manufacturer and I decide to do this myself, it sounds like there is going to be a lot of investment in hardware, software to collect data and then contextualize the data. I don't want to minimize that point, either, because contextualization is so incredibly important. It can collect an awful lot, but if you don't know what the number means and put it in something that gives it a reference, it's hard to make a decision based on it. So, are there vendors out there, are there activities out there, are there organizations out there that are taking and pulling this all together under one roof and making it more interactable? How does the industry need to evolve so that interoperability and simplicity come into place for the end-user, where they don't have to spend 90 percent of the cost or 95 percent of the cost just getting ready to get ready? I think that's what I heard you describe.
Jim Davis [00:19:20] So, Gregg, if I can, let me take you through a couple of dimensions or aspects to that question. I think it's important to unpack it a little bit. One of the things where we can take a lesson — and I mentioned this just a moment ago on the IT industry — the IT industry has been very, very effective at scaling. We're very, very used to that. All you have to do is think about how I use the iPhone, how do I get to advertisements, think about consumer products, think about the banking industry, these sorts of things. Those kinds of scaled aspects of using the IT have not found their way into manufacturing, and there's a number of reasons. If you step back... I'm going to play off your comment about the contextualization of the data. If you really step back and say, "What do I need to do with that data? Where does it come from?" you'll very, very quickly come to the conclusion that number one is I have to generate it, I have to have a source for it, and it's usually a sensor or human entered. I then need to connect it. I then need to provision it. So, that's where contextualization, labeling, tagging, a variety of things that bring that data together into the form that you need to use it. I need to orchestrate it. So, I often need to put it together with a model or maybe different models, or I have to put it together with the process. So, workflow comes into play very, very strongly just with using data. Then I have to put the data, the workflow, the models all together. So, you have to execute with it in some way, shape, or form, and you have to execute to purpose. Am I controlling? Am I diagnosing? Am I monitoring? What am I doing with it? Then an important aspect that doesn't happen so often in manufacturing is you want to archive those results. You want to archive the data, and you want to archive the capability so it can be reused. That archiving piece becomes very, very important. That's what happens in the IT space, but let's talk about it from a manufacturing space. So, again, sensors and data come from humans. So, how do I reproduce that? How can I learn from that and scale from that? I have to ingest it. I have to store it someplace. Do I have to store it in my own storage all the time, or can I work with the vendors on some storage that meets those specifications? The contextualization is a heavy lift. I have to know my process. I have to know the domain. But do I need to do it every single time, and does every manufacturer need to do it every single time for the same machine? That's an important question. The workflow — do I have to build the workflow the same way the same time for the same machine or the same process if they're used commonly across the industry? I'm not talking about getting into the intellectual property at all. Then how do I use the models? How do I use the software out there? Can I scale that? What are the best ways to do that? Then how do I put them in a form using the IT tools to reuse it? So, you can see these parallel questions between IT and OT. This is what we're really spending a lot of time on is how to scale and reuse across each of those steps with data. So, that's one dimension of this.
Gregg Profozich [00:22:29] So, Smart is essentially bringing some of these scaling processes in ways that IT has matured over the past 20 or 30 years and beginning to apply it to the IT/OT bridge. It becomes I'm a manufacturer; I have equipment on my production floor; I don't have good data off of it yet. When I can get data off of it in context and use it to make good decisions, then I'm really leveraging the true capability of what I need to do to make improvements and run that way. So, Smart has tools that become the bridge between IT and OT. Is that what we're saying?
Jim Davis [00:23:02] We are definitely saying that. There's lessons learned in other industries which have been able to scale and reuse the IT, moved into interoperability, data exchange, and so forth. But what I was trying to draw is a parallel between the IT and the OT. There's so much out there about this OT/IT integration. It's another one of these buzz terms that is being used in many, many different ways. So, I'm really just describing how I use what the integration means. It's really tying the operations directly with the IT so that we can scale and use the same kinds of capabilities in the manufacturing space. There's still some challenges, which we'll get to in a minute.
Gregg Profozich [00:23:42] Got it. So, with all this information in hand now, I think we've got a really good context and a handle on a lot of the key concepts that you laid out for us. Can you tell us about how broadly Smart Manufacturing has been adopted and share some barriers that you see to even broader adoption?
Jim Davis [00:23:57] So, I've been looking at this for 10 years. The adoption is not very far. It's not very far along at all. One of the reasons is that the whole transformation in what we call Smart Manufacturing, or the Fourth Industrial Revolution, and so forth, is still in early stages. But the challenges are really worth looking at because they have to do with the third generation and what needs to change there from a business practice. The third generation is exactly what you said. We've moved into the digitization — not the digitalization, but the digitization — of manufacturing and manufacturing operations. If you go back far enough, we've moved from analog processes, analog sensors, and controllers into digital controllers, but we really didn't change how we did things; we really didn't change the objectives. So, my term on that is digitization. If you really pull back the cover on how manufacturing looks at things, the focus is still heavily on the assets. It's still heavily on the function; it's still heavily on how I'm manufacturing things, how have I structured things. So, it typically is engineered to order. It could be made to order; it could be made to stock. It could be I have a discrete process; it could be I have a batch process, and I have a continuous process. So, when you write software that's focused on engineered to order for batch processes, what you've done is you've focused on the function and the assets, and you've buried the data. So, you don't have a good way of extracting the data; you don't have a good way of having contextualized that data so you can reuse it. So, you've in effect trapped the data in the software in the ways you're using it. So, one of the things you want to do from a Smart standpoint is untrap the data. Another dimension is the one I mentioned already. If we keep building our own IT infrastructure, data management systems, and so forth, you're actually spending 90 percent on a particular system that doesn't even get to the value yet. So, we have to lower that cost. This is where this term pilot purgatory comes from. Small companies have trouble even getting started; large companies keep building out a pilot. Then they find it's very expensive because they're in their own infrastructure, they can't scale it very easily. Then it just falls apart. There's a vendor component, Gregg, which you were alluding to. We have the age of cloud now. The vendor market, appropriately so, has grown up around apps and infrastructure combined. So, it really goes down this path of buy my application and, therefore, buy my infrastructure, but what we really want to do is interoperate with different tools; we want to interoperate with different companies; we want to interoperate across line operations. So, the whole vendor market is actually structured against what we're trying to do from Smart Manufacturing. To be honest, cloud has made it worse. It solved the infrastructure problem, but it's made the integration and the interoperability just as hard, if not harder, because the vendor market goes down the path of buy my cloud product, use my cloud product, and that's not what's useful in our particular case.
Gregg Profozich [00:27:20] It sounds like we're ending up with a lot of individual suboptimizations in silos.
Jim Davis [00:27:25] We're optimizing for vertical instead of horizontal, which is where the benefits are in Smart Manufacturing. We did a little benchmarking study. You bring a software product in, and you say, "What does it take to actually configure that for my process, and what does it take to configure and integrate it into a secure environment?" We were seeing on the order of 800 to over 1,000 different parameter settings, configuration settings, to set up one single piece of software. The software products tend to have quite a few possibilities, bells and whistles, and so forth. So, they're not simple is the point. A typical operation, even a small company operation, probably will want to use a number of these, and a large company we'll see wanting to use 50, 80, 100 of these kinds of products, everywhere from the financial to the inventory management to the maintenance management to the operations and control and so forth. It's those kinds of things. So, if you think about Smart Manufacturing and wanting to do more and more with data and more and more with these applications, you're increasing complexity, because you're increasing the need to do all this configuration and setup on these kinds of things, it becomes impossible. Then the other thing that's desirable from a Smart Manufacturing standpoint is to exchange data between the two products. We did a little benchmark, and it takes about 250 person-hours to actually set up a secure interconnection between the two within whatever IT environment happens to be in vogue or in practice with respect to that particular company. So, you start multiplying interconnections between these, the problem becomes more and more complex. There's a bottom-line statement here from my perspective. If the industry keeps going the way it is, we're managing to the greatest complexity, not to the greatest simplicity. We have to get around that.
Gregg Profozich [00:29:20] Until we simplify it, it's going to be an issue for broader adoption. So, we've talked a little bit about the current state. Can you give us a little preview, or review, actually, of Smart Manufacturing and how it's developed over the past 10 years? What are the key technological developments that have enabled Smart to become viable that our listeners should be aware of? Is it the cloud? Is it analytics? Is it sensor technology? What are the things that are critical?
Jim Davis [00:29:44] The thing that's out there right now is significant capabilities to manage the data. These, again, are heavily drawing upon the IT learning, but the cloud is very, very important. We just need to make it interoperable. One of the big areas in vogue from a vocabulary standpoint right now is under the heading of artificial intelligence and machine learning. I know it's got a bad rep out there because the rep is we're teaching computers how to think like humans, and then we're going to replace humans. From my perspective, it's really not the case. It's really a set of technologies that now allow us to manage the data, contextualize the data, do analytics that are just unprecedented from what we could do before, ways to pull data together to do the things that we're talking about with Smart Manufacturing. So, it needs the domain expertise; it needs the knowledge and know-how of the particular manufacturing processes. But the tools and capability, cloud, the ability to exchange data, to exchange it securely — which is another aspect we need to talk about — are all out there and, I think, significantly in place. So, what we're really talking about is getting started and changing some of the business practices, changing some of the vendor practices so that we can move forward with this horizontalization is my term that I use here instead of vertical orientation with respect to the build-out of manufacturing. So, there's a lot of capability out there. I'm really quite excited because there's so much understanding and capability that's now starting to take hold, shall we say, 20 years, 25 years after the Web has started, which is pretty fast.
Gregg Profozich [00:31:29] Yeah. I think the landline phone took something like 35 years to get to 25 percent adoption across the nation. The Internet is worldwide at this point in less than 25. That's pretty stunning. The pace of innovation is definitely speeding up. So, you mentioned a little bit about security issues and cybersecurity. I think there's some things that you would like to share with us about that in the context of this conversation. So, what are the key things in the cyber world and the key things to be aware of when implementing Smart to make sure that you've got a secure environment and a secure solution?
Jim Davis [00:31:59] On the security side there's quite a bit to go into. Let me just roll it up to a couple of principles from a Smart Manufacturing standpoint. The first thing is it's important, and you need to do it. It literally starts with the sensors, and it starts with the data, and so forth. One of the things that we're seeing with respect to Smart Manufacturing is that the focus on the data now is giving each manufacturer much more insight into the data, how it's being used, what's inventoried, those sorts of things, which are important basic first steps with respect to the security. The second thing is that the cloud is an important security tool. You need to think very, very, carefully about using that. You have to think and understand the role of the cloud, and how that plays, and what the role of the manufacturer is. But nevertheless, the cloud is an important security measure. One of the things that we are concerned about from a Smart Manufacturing standpoint is that this proliferation of interconnections, this proliferation of trying to do interoperability across different infrastructures, and different companies trying to do it on their own. What that does is proliferating one-off infrastructures and trying to deal with security in one-off manners. That's a threat. So, working as an industry and working carefully on this interoperability question, and how we do and exchange data and so forth in a much more consistent manner with understood standards and practices and so forth, becomes very important. If we increase the complexity of manufacturing, we're going to increase the security threat.
Gregg Profozich [00:33:41] All right. So, in terms of broader adoption, technologies seem to be broadly and rapidly adopted when they help solve real problems in an effective yet simple and user-friendly manner. Think about a smartphone application. I really don't need training to access and use turn-by-turn directions to get from point A to point B these days. So, where are we on the effective and simple user-friendly continuums as related to Smart Manufacturing? Where do you see it getting down to more of a download the app and plug-and-play world?
Jim Davis [00:34:13] So, we're at the very first stages, but it's real, and it's out there, and it is starting to be used. That question actually moves us into how can you begin to think about scalability and reusability. That is the job of the Manufacturing USA Institutes and a number of members going forward, a number of different coalitions that are working on this together. This is the role of public/private partnerships. I'm a very strong advocate of those. But there needs to be a way for the industry to work together — we call it industry-wide strategies — and the ability to put together infrastructure in a consistent way with the vendors that allows for this connectivity, this provisioning, the orchestration, the execution, workflow is very much in place. Then the ability to manage the data, contextualize it, and build out reusable we call them templates or configurations for common uses, common machines, common processes. That's all well underway. This is all moving us towards a scalability and reusability. These are available now for small, medium, and large companies to try out, to kick the tires, that sort of thing going forward. I'm the first one to say that there's work to be done, but the work is going on, and moving in, I think, a very, very appropriate direction to get at the headroom of economic benefit, safety, security, all the things that we've been talking about.
Gregg Profozich [00:35:44] So, the vision sounds a lot to me like if I'm a bakery and I've got a series of ovens that I use, if there was a standard model, pull in this data, put it in, here's a configurator and a contextualization engine, and here's a calculation engine, it becomes more plug and play. Is that the vision and the direction we're trying to go here with Smart, where we don't have to do a custom set for every oven, but we have categories or ovens by manufacturer or something along those lines?
Jim Davis [00:36:12] That's exactly what we're trying to do. So, if you take the oven example, there could be different kinds of ovens, different configurations, but that's a smaller subset. So, if you knew where to put a thermocouple, or if you knew where to put an infrared sensor, those sorts of things, how could that be installed? Perhaps a manufacturer does that already. But then the question becomes: how do you connect that sensor data? So, if you had a sensor connection box that collected that data in a standard configuration, and put that data on the Web securely, or on your network securely, wherever it needs to go, if that could then be tied in, as you just said, to an analytical product, dashboard product, can be tied into a controller of some sort that could come through the vendor of that oven, the IT infrastructure could come through aggregated products through the cloud, through vendors, through these institutes that we're talking about. So, all of this can be simplified. We believe that the IT environment, the cost could be dropped at least in half going forward, and connecting this up could be dropped significantly. That's going to have a significant bearing on that ROI and make this much more reachable going forward. So, yeah, all of these components are readily scalable, and we're just working to do that within the industry.
Gregg Profozich [00:37:37] Okay. So, Jim, I'm gathering from our conversation that Smart Manufacturing is a very data-centric domain. In your opinion, what are both the key cultural elements and the technical capabilities that an organization needs to have in place in order to get the full value from an investment in Smart?
Jim Davis [00:37:53] What we're seeing in the manufacturing industry on the cultural side really has to do with risk. There's a fear of the data. There's a fear of changing things. There's a fear of working with serviceable equipment, and making that data available, or connecting that up, those sorts of things. So, just the risk of going down this whole path of Smart Manufacturing and dealing with the IT, all the sorts of things that we've talked about. The intellectual property — I've mentioned that a couple of times. The manufacturing industry holds the intellectual property to a very, very high degree. Anything that has to do with the process, anything having to do with the product, anything having to do with virtually any kind of data about the process or the product is considered to be intellectual property and not released. I cannot release it is the general mode of operation. What we're seeing is that if the culture can change and recalibrate what intellectual property is, especially around these common machines and common processes, you actually can abstract and separate the secret sauce intellectual property from the operational intellectual property that becomes very useful for scaling. The crawl, walk, run point that was brought up earlier is extremely important. Don't spend a lot of time as your first project putting in lots of different sensors and a really complicated IT infrastructure. It's far, far better to start with one or two sensors; have a very, very specific objective; keep the IT infrastructure as simple as possible. There is infrastructure out there to work with. Then get used to thinking about the data. Get used to thinking about all the things you have to do to the data just to get it ready to interface with a model or be used by a model going forward. Understand the tools to do that. Those practices, that understanding are extremely important. Then that obviously gets to the workforce side of things. Workforce training is extremely important. But we will also say that having the tools in the hands of the workforce to do the training and to move into some kind of real problem all at the same time is a really important alignment. It doesn't work so well just to train and then come back a year later with some kind of problem and so forth. Let's get those aligned. Starting small facilitates that, as well.
Gregg Profozich [00:40:20] So, Jim, what do you see is the next evolution of Smart Manufacturing in terms of technology and trends?
Jim Davis [00:40:25] Gregg, let me first of all talk about the macroeconomics a bit. We've done some economic studies across many different industry segments. These range paper, steel, metal, glass, food, electronics, oil and gas coatings, plastics — these aren't necessarily defined segments, but their areas — automotive, aerospace. What we're seeing from an economic value standpoint is the starting point is 15 percent economic value, and it goes up from there if you can get started with this technology. It goes back to those three terms: productivity, precision, and performance. It's not about driving the efficiency of any individual machine or operation. It's about reducing waste; it's about increasing productivity; it's about the value of increased precision; it's about the value of increased agility, and overall performance, and those kinds of terms. That's where the data becomes important. Our economic study showed that if we had just 10 percent adoption across small, medium, and large companies, it's not a cost-cutting situation; it's how to stimulate an investment. Ten percent adoption leads to one and a half million new jobs or indirect jobs. Granted, they change character, but the job market actually increases on this. The investment potential is $200 billion over a number of years. That's huge. It's not cost-cutting. So, you really want to get going on this to get ready for the investment side of it and the potential it has for you as a company and as employees working within those companies. So, this circles back to the starting point. You have to get started with that single sensor, and thinking about the data, and working with those practices, and so forth. We see that the economic value is going to start generally around quality. We see quality assurance in some way, shape, or form being the number one focus and then the prediction around maintenance, predicting capabilities around processes and machines. Those two stand out significantly on the factory floor beginning down that path of the sensor. Then how one can scale that and scale the modeling and the capabilities around those two areas, I think, becomes very important. What we see is if you can take on that single sensor and the practice, you can't stop there. You have to move into what we call industry-wide strategies. So, if you can take that data and now think about work and that data add value to my company by working with another company, if I work with another company or I work with a vendor, where's the benefit come to me, where's the benefit come to the country? There's a number of places in which that shows up. So, it's the data interoperability, and it is the ability to exchange data to build out some of these models and these configurations around these common processes. So, all I'm really saying is start on the factory floor, start simple, and then start thinking about how you can work more broadly within the industry because that sets up these industry-wide strategies.
Gregg Profozich [00:43:36] These evolutions that are happening, it's a rapidly changing environment. What should I be doing to best prepare myself for the future that's going to be here soon?
Jim Davis [00:43:44] I think the best thing is to work with those organizations now that are taking this viewpoint on manufacturing. So, I've already mentioned the Manufacturing USA Institutes. They are taking this forward. The Clean Energy Smart Manufacturing Innovation Institute, which I'm affiliated with, is certainly doing this. But it's not just the institutes. Those institutes are working with the industrial assessment centers, with the Department of Energy. Those are important to build out these kinds of practices. Gregg, as you know, I'm a very strong advocate of the MEP programs and the training and the consultation that’s brought forward there. The MEP programs are taking these kinds of technologies forward. They're building in really important security measures, really important energy measures. Smart Manufacturing integrated within those kinds of consultations, and training, and so forth are really important to take this forward. It'll build out of the capability that's going to be needed for both the manufacturing floor, the factory floor, and these industry-wide strategies that I'm talking about. The major thing I'm saying is that there are organizations out there that are taking this on, and the biggest mistake would be to not tap into those and start building out that practice and understanding about how it's going to apply to my factory, how it's going to be applied to my machines, or whatever I'm doing.
Gregg Profozich [00:45:09] Okay. Excellent. Any closing thoughts before we wrap up?
Jim Davis [00:45:12] One other thing I might just add, just to close with the definition of Smart Manufacturing. From an institute standpoint, I know I started out with the right data at the right time in the right place, and so forth. If you want to look at it, there's another set of rights. It is really about the right product made the right way the first time. So, that set of rights becomes very important. If you really want to do that, you have to get beyond just the factory machines, a factory floor. That's where Smart Manufacturing plays a significant role. So, the key properties, in summary, for us from a Smart Manufacturing standpoint are interoperability — or I use the term horizontalization, not just verticalization — it is security; it is scalability; it's real-time. This all has to operate in time. So, decisions, actions, dashboards, control systems are making decisions, taking actions in the right time. Really important is to be able to move into the predictive space. There's a lot of analytics, a lot of monitoring, diagnostics, and control which are important, but the holy grail is to be able to predict. That's where the IT, the OT, and Smart Manufacturing take you. Think about the operations in the light of the pandemic. So, the notion of resilience and orchestration become very, very important for everyone, and then the environment and the sustainability environment. Those are all key elements of Smart Manufacturing.
Gregg Profozich [00:46:50] Excellent. Great information. So, if I was to try to summarize my key takeaways from our conversation, the litany of rights is one of the first things. How do I get the right data to the right place in the hands of the right people at the right time across a single organization or across an integrated set of supply chain partners, an integrated enterprise? I think that crawl, walk, run approach is a key thing to keep in mind. Start with a single sensor. If Smart Manufacturing is using a sensor and monitoring communications via Bluetooth, or wireless, or hardwired to get information to the cloud, potentially, to be able to do analytics on it and provide it back in dashboards, that's an end-to-end look at one of the ways that Smart can be implemented. As the technology develops and matures, there's going to be much more interoperability across supply chains and the ability for multiple organizations to work as a single entity, if you will. I have my production line here in California; my next trading partner downstream is in Arizona; the next one's in Kentucky. Well, the three of us could be sharing information real-time in live and making real-time adjustments that way. Smart enables higher productivity, higher precision, and higher performance. The pressures out there in the industry are for higher precision, smaller lots, faster and cheaper, track and traceability. All of these things can be enabled by Smart. Smart is really about the data, unlocking the data, and managing the data. That can be done through individuals in the workforce who have those skills. It can also be done through the application of artificial intelligence and machine learning. All of this has to be wrapped up within a secure environment. So, the principles of cyber, making sure you do it and making sure that you have security built-in and use the cloud for the built-in security features will help manufacturers to manage risk and intellectual property. That about sums it up, I think, Did I miss anything in there or anything you'd say differently?
Jim Davis [00:48:40] Yes. So, Gregg, your summary is, I think, right on, and I'll roll it up one step higher. You have to look beyond just the factory. Industry-wide strategies are going to happen. They have to happen. Industry-wide strategies will impact the infrastructure, it will make it better, give you more capability. But the industry-wide strategy will be critical for the business because they will move us down the path of the interoperability and the ways to leverage and scale the use of this data and capability. Small, medium, and large manufacturers alike all need to have this capability for the industry to be at its highest achievement level with respect to Smart Manufacturing.
Gregg Profozich [00:49:21] So, Jim, I really want to thank you for joining me today and for sharing your perspectives and insights with me and our listeners. There was a lot of great information and some great takeaways, I think, in this conversation. So, thank you for that.
Jim Davis [00:49:32] Gregg, it's always a pleasure to talk with you. I know you and I have talked about this for a long time. As we said at the beginning, going back six years. I'm such a strong advocate for the MEP programs. It's such a pleasure to be part of this information going to your programs. Very, very happy to be here today.
Gregg Profozich [00:49:47] To our listeners, thank you for joining us for this conversation with Dr. Jim Davis on Smart Manufacturing Smart for your business. Thank you so much for being here today. Have a great day. Stay safe and healthy. Thank you for listening to Shifting Gears — a podcast from CMTC. If you enjoyed this episode, please share it with others and post it on your social media platforms. You can subscribe to our podcast on Apple Podcast, Spotify, or your preferred podcast directory. For more information on our topic, please visit www.cmtc.com/shiftinggears. CMTC is a private nonprofit organization that provides technical assistance, workforce development, and consulting services to small- and medium-sized manufacturers throughout the state of California. CMTC's mission is to serve as a trusted advisor, providing solutions that increase the productivity and competitiveness of California's manufacturers. CMTC operates under a cooperative agreement for the state of California with the Hollings Manufacturing Extension Partnership Program (MEP) at the National Institutes of Standards and Technology within the Department of Commerce. For more information about CMTC please visit www.cmtc.com. For more information about the MEP National Network, or to find your local MEP center visit www.nist.gov/mep.
