Erik: Welcome to the Industrial IoT Spotlight, your number one spot for insight from industrial IoT thought leaders who are transforming businesses today with your host, Erik Walenza.
Welcome back to the Industrial IoT Spotlight podcast. I'm your host, Erik Walenza, CEO of IoT ONE, the consultancy that helps companies create value from data to achieve growth. And our guest today will be Keith Higgin's, Vice President of Digital Transformation and Industrial IoT at Rockwell Automation. Rockwell brings edge to enterprise analytics, machine learning, industrial Internet of Things and augmented reality into industrial operations. In this talk, we discussed how to improve how your company accesses, processes and leverages data to make better decisions. We also explored four use cases: scalable production management, enterprise operational intelligence, intelligent asset optimization, and digital workforce productivity.
If you find these conversations valuable, please leave us a comment and a five-star review. And if you'd like to share your company's story or recommend a speaker, please email us at team@IoTone.com. Thank you. Keith, thank you so much for joining us on the podcast today.
Keith: Thanks for having us here. Really appreciate it.
Erik: But before we dive into the topic, and how you approach it, I want to just introduce to our audience who you are, and how you ended up as VP of Marketing Digital Transformation for IoT over at Rockwell. I saw that you most recently were with Foghorn which is quite a prominent startup in the space. You've worked in senior positions for quite a number of high growth VC backed companies. What was it that led you now to your current role with Rockwell?
Keith: Well, what I realized about the digital transformation space is that so much of the success of those programs, those initiatives, digital twins, industrial IoT gateways, predictive maintenance, they're really predicated on understanding the context of the OT data, of really what is that data telling you, where's it coming from, and how do I get that into a digital system. And Rockwell with 100 years of operating as a supplier to some of the world's leading industrial organizations has a really good understanding of everything that's happened to date in the industrial segment. And I think that gives them a rare opportunity to deliver the next generation of digital solutions with a real grounding in OT data. And I think in all the time I've spent in my career in networking and now in this space of industrial IOT and digital transformation, that's the one single theme that really brought me here.
Erik: So this topic of OT IoT connectivity is going to be super relevant here. But before we dive into that topic, which I think we can go into a fair amount of depth into, maybe we can touch first on the types of companies that you're working with at Rockwell, and what are the challenges that they're facing? So why is this now a priority for companies in oil and gas, companies in pharma in food and beverage? What is it now that makes this topic a high priority?
Keith: Well, in some ways, it's more of an opportunity than a challenge. I think what's happened now is that the early adopters of digital transformation are getting products to market faster, they're reducing their operating costs, they're getting a lot more efficiency out of their assets. So there's a lot of people that believe or moving into a world where digital transformation is required to be competitive. It's somewhat the new competitive landscape.
And there's a lot of studies out there talking about the rapid opportunity that opens when industries become digital, and the risk from very large firms to be displaced by new technologies and companies moving a little bit quicker, a little bit faster. So I think those are some of the opportunities in terms of really getting people products to market faster, operating at the highest levels of efficiency, getting the most out of their operations, being able to fix things before they break, all the positives you've heard of these, what I would call lighthouse implementations of industrial IOT and digital transformation.
Erik: This is kind of it's still somewhat of an open question, for me at least, which is over the past 30-40 years in a lot of industries, in media, in finance, and entertainment, you've seen the incumbent companies move a bit too slowly and adopting new technologies, and you've seen new players really come to the fore and take a very significant market share. In a lot of the traditional industries, I'd say you haven't really seen that yet because I think it's more difficult to scale quickly in these industries, but we might begin to see that now if the incumbents aren't able to, to evolve quickly.
How do you see that? Do you think that the digitalization of traditional industries, heavy industries that have big assets and asset basis, do you see them also as being open to disruption in the same way that the media companies of the 20th century were open to disruption? Or do you think that because they have this physical presence in the world, it's fundamentally different in terms of how susceptible they are to change from new players?
Keith: I think there's great susceptibility. There's great examples in the industrial organizations, where people now are coming in and doing mass customization producing things faster, much lower cost. I think I was watching a mattress commercial last night with one of my kids, I was like how is somebody disrupting that industry? So, I think it's a combination of the technologies that we've seen happen in the retail industry in other areas combined with the industrial world where we are going to see some disruption. And I think in some ways, we've already seen that.
So the good news is I think these tools now are being deployed at significant scale. And there's a lot of very well documented high value use cases for digital transformation for companies to really improve their yield really lower their costs, really improve their productivity, things like that.
Erik: I think this concept of digital transformation it's kind of like the elephant and the blind men, it means a lot of different things to different people, depending on where you sit in the organization. What does digital transformation mean for Rockwell? And what does it mean for the customers that you're working with?
Keith: Well, for us digital transformation really is about turning OT data into actionable insights, and using that to improve the financial performance of a company, and to improve the operational efficiencies. And four of the predominant high value use cases that have emerged in digital transformation largely are around what I would call intelligent asset optimization, which is how do I take all these really expensive pieces of equipment that I own and make sure I'm getting the most out of them and they last the longest, and they produce the most? Another one is how do I look at my entire operation and have real time analytics, so enterprise operational intelligence is a big area of focus.
A third or fourth, which is really kind of scalable production management, which is how do I have great granularity performance of my manufacturing line all the way back to track and trace of inventory and different pieces. And one of the most exciting areas I think that people are hearing a lot about now is in this area of digital workforce productivity, and how do I use technologies like augmented reality and things like that to transfer knowledge faster and to address one of the biggest challenges you hear a lot about in the industrial organization, which is the aging of that workforce?
Erik: If we look at these fours, so we have intelligent asset optimization, enterprise intelligence, scalable production management, and then digital workforce productivity, how would you look at the maturity of the solutions here today, maybe not just from Rockwell’s perspective, but across the market? Because I think what we've seen in the past five years or so is a fairly rapid improvement in the maturity of solutions.
So companies going from pilots and a lot of uncertainty about what value can be created, what makes sense for them towards now environment that for me feels like we have much better data around what ROI might be, what potential challenges might be in terms of during deployment and so forth. But we're certainly still not at a situation where these technologies would be considered mature like the internal combustion engine. How would you look at these four use cases in terms of their maturity and ability to scale in a more standardized almost commoditize way? Are we still looking at very customized deployments or do we have solutions that are really almost plug and play to some degree at least?
Keith: Well, if you're going to compare these solutions to the internal combustion engine, I think we've got a ways to go. I do believe that the maturity of the solutions today is very high. I personally, think about the digital transformation segments of customers. One is, are they collecting data today? So they have a historian, they're thinking about, hey, I want to try to monetize my data. The second phase, which I think a lot of customers are moving into, which is, how do I visualize that? How do I take all this disparate data from these different sources? And how do I visualize it, do some analytics on it, and start to get some value and intelligence and insight into the data being generated from my operations? I think the third progression of that is what we would call a system of record, which is a much more granular tracking and view of real time production performance, and all the way tied into ERP, tied into many of the other elements that are around that?
And then the last stage is what I would call the real kind of advanced companies who have taken this concept of machine learning, artificial intelligence, advanced analytics, and roll that out to multiple sites. So I think you're right, there's definitely a maturity curve in terms of capabilities, were probably the long tails in the data visualization stage, or they know there's value in there, they’re trying to find it in those high value use cases. The middle of that market has moved to a system of record and production management environment with a lot of data and a lot of insights. And then you have the tip of this market, which is really adopting quickly, everything they can to get the most out of the operational data that's being generated.
Erik: So we might be looking at 50% of the market, that's still kind of in the data collecting stage, maybe 30% that's visualization, 15% that’s at the system of records stage and maybe the 5% of companies that are really doing now advanced deployment. You kind of look at it like a pyramid like that, or are we talking about still 85% of companies that are still at the data collecting stage?
Keith: I think it's closer to the former than the latter. But I certainly think it's like I say closer to the former than the latter, which is more companies have moved to this type of approach. I think there was a statistic recently that 90% of industrial companies now are investing in some form of digital transformation. And so, as that investment cycle becomes more and more, I think, we'll see those companies move through that process more. But I think your first stab there is probably pretty close.
Erik: What do you see as the big challenges as a company progresses through these phases? Are we talking about just the will to put the capital into play to make the investments in hardware and software? Are we talking about companies still figuring out what's the right approach for them? Is this more of figuring out what strategy makes sense, given their objectives? When you're advising companies that are maybe at the earlier stages that are still collecting data, still figuring out how to visualize, what seems to you to be the primary challenges that keep them from maybe progressing as quickly as they would like to?
Keith: I had to narrow it down, I think for many customers it's really understanding where the highest value use cases reside. It starts with what are those use cases that we think there's a significant payback to sort of attack. And then are we aligned organizationally? Are we aligned with a digital transformation office? Do we have the OT-IT sides of the fence together on this? And do we have targets that we're aiming for as we start the initial rollouts and then the full scale production. So, those tend to be the pieces, the use cases, the team and the budget and are those aligned with the strategy of the company to move down this path. And increasingly, this seems to be the new competitive landscape in the industrial space. So more and more companies are not now saying if, but they're saying when, how and how fast.
Erik: Why don't we dive now into Rockwell's approach? So what would you say differentiates Rockwell from other providers in the market in terms of your approach for digital transformation?
Keith: Well, I think the number one thing that differentiates us is the fact that we've spent 100 years operating in the industrial environment close with our customers, understanding how machines work, how data is collected, how production lines work. And we've starting several years ago began to invest heavily in how those industrial environments and that OT data can be leveraged best to deliver financial results.
And one of the elements of that strategy was the billion dollar investment for made in PTC, maybe a year and a half ago, now. And PTC, as you may know, is a leader in digital transformation. They have an industrial IoT platform called ThingWorx. We've taken that technology along with our factory talk platform, which is our suite of industrial edge software. And we've combined those two solutions with heavy development and integration work into a solution that we call Innovation Suite. And to my knowledge, it's really the first software suite designed for the industrial edge to bring together the OT world with all of its contexts and all of the deep meaning that happens when OT data is generated, and marry that with the digital transformation world, the industrial IoT platform world in a way that you have no erosion of the value of the data. And that's the single biggest differentiation of our solution.
Erik: Okay, the name itself Innovation Suite is somewhat interesting. Because I think if you look back 10 or 20 years, when you're talking about industrial software, and so forth, the word ‘Innovation’ wouldn't really be in there unless it was maybe, you know, software for the R&D function. But if you're selling into oil and gas platform, you're not going to put that word in right then. Maybe you can talk to us through a deployment of Innovation Suite. So what would this actually mean for customer in terms of the data that they would be ingesting into the system, how they would be extracting value from that, and who would be accessing the system and so forth? So what does this look like if we use a standard deployment?
Keith: Well, I think standard deployment is I've got an existing operational environment and I want to get the most insight and value out of that data and I would work with a company like Rockwell to look at the various, what I would call modular and industry tailored offerings within Innovation Suite. So I would select something that was right sized for me in terms of my operations and I would select something that was vertically tailored to the space that I'm in. And I would start to look at back to those three pillars of digital transformation. I want to start to visualize my data, I want to move to a detailed system of record, or I want to move to a complete full advanced analytics environment.
And I think what's most interesting right now is the integration, the four straight lines, if you will, between OT and IT. And when you start to breeze worlds together, the first challenge you have is discovery. Like how do I know what OT infrastructure that I have today? And how do I make that as easy as possible for a new digital transformation solution or industrial IoT platform to discover? So that's a big technical challenge, I think, to solve that delivers a lot of value to the industry today.
The next example is how do I keep all of that OT data special context? Someone used analogy the other day that I thought was interesting, which is, if you're in a conference room, there might be a temperature sensor. But telling the temperature is one thing. But if you map it to the time of day, is the sun out? Is it shining on the window? How many people are in the room? What type of electronic equipment is operating? Like there's a lot of context to the data and to the situation. And how does all of that type of context be captured and pass seamlessly without, again, erosion into these different IT initiatives, and digital transformation initiatives?
And then the third piece is how do I start to look at analytical mashups between that contextual data I'm getting out of the OT environment, and what these new analytical tools are providing me in the industrial IoT area, for example, our partner PTC with ThingWorx? And finally, again, one of the areas is how do I start to connect this OT data effectively with things like augmented reality? So how do I allow a newly hired worker to perform maintenance on a machine by seeing augmented reality work instructions in front of them in a screen, perhaps with that same procedure having been performed by a more experienced technician? How does that make people productive faster? How does it close the workforce gap?
And those are all things by the way that we do today, to answer your initial question, I hope that separates Innovation Suite as a solution in this market right now, between Rockwell and PTC. We've done quite a bit of quantification of that. We believe we have a significant differentiation and market lead in terms of production environments today.
Erik: So just to quickly summarize to make sure I have this straight, so we're looking at discovery of data, and ingested and contextualization of the data, making sure you have the metadata structured properly, then the analytical mash up so that people, whether it's management, or maybe a maintenance engineer, can make use of that data. And then we have connecting the OT and the data with other solutions. Maybe AR probably could be, I suppose maybe also, 3D printing or other types of solutions that somebody might want to deploy.
This, I think, for a lot of companies it's complicated, because both the hardware that they have deployed, whether it's sensors, or the equipment in their facilities, or even the software might be decades old. How do you approach the topic of future proofing a solution? So somebody doesn't end up making an investment in 2020? And then two years later, realize that AR has evolved quickly and that there hasn't been some change that then requires a significant modification in that architecture?
Keith: Yeah, you've hit a really important point, which is that there is a lot of legacy infrastructure. So those protocols, everything needs to be unified through a data ingestion layer. And that's one thing that's offered with Innovation Suite is the ability to ingest the great majority of legacy protocols that are in use today and to unify that data. I think having a platform approach that is designed for interoperability with what I would call best in class ecosystem solutions, because those will change over time, you will get customer preferences and cloud, you'll get customer preferences in simulation.
What we're doing with Innovation Suite is not limited to Rockwell hardware. It is a solution that can be deployed in environments that is to some extent, hardware agnostic. So all of those pieces are important to the future proof point that you bring up, which is how do I know what I am bringing in is foundational is designed for change, both ingress and egress, and to some extent, sideways, if you will, depending on how you've mapped your industrial IoT architecture? And those are all things that can be really well tested with the vendors that you work with by asking that exact question.
Erik: And you just mentioned one point there that has been a big question mark for a lot of the companies that we work with, which is when you're adopting a platform or building a platform internally, the cloud, so the public versus private versus on-premise question, especially when you're talking about manufacturing environments, or these industrial environments where typically things have been pretty much on-premise, and companies are pretty conservative, what's the viewpoint now for factory talk in Innovation Suite? Do you see a significant shift towards using cloud? And if so, is it more public or private? Or do the majority of companies still prefer to keep the data on-premise and maybe only put some small amount of the data onto the cloud?
Keith: Well, I think in the early stages of things like operational analytics, cloud was maybe some of the earlier approaches. And we are very much in partnership with companies like Microsoft for the cloud elements of this. I think they're very complementary edge and cloud.
But if you look at what's really happening at the edge, with data volume, things like video sensors, for example, there's simply too much data to send all of it to the cloud. There's security implications. There's latency implications. There's cost implications, communications, infrastructure implications. I think this is why Gartner and other analysts have come out and said we're going to shift to a world where quite a bit of data is processed at the edge for all the reasons that we just talked about.
And then the question is what is the edge to cloud architecture that optimizes the business case for digital transformation? And I think you'll see a lot more from Rockwell, as we work with partners like Microsoft on what is that reference architecture look like? What are the benefits? And how does that really make it a lot faster time to value for customers as they implement these digital initiatives?
Erik: So we're moving to a world where a lot of the work is being done on the edge and then companies need to decide what is going to be sent to the cloud so that it can be shared across the enterprise?
Keith: Yeah, I'd be careful to say work at the edge because then it sounds like less work at the cloud. It's not really less work at the cloud and more work at the edge. It's more of a rebalancing. For so long, we've lived in a cloud centric world, because we haven't had new software and capabilities that do magic closer to the data source. So I think we're going to move to a world where we have those specialized software capabilities that were designed specifically for the industrial edge close to the data source, close to the systems that are generating the data. And that brings out the best in the cloud, in terms of what it's really good at, things like machine learning, model training, a lot of the heavy data work, if you will, I think it's going to be a really good balance.
The thing I'm excited about is, to me, I say this kind of tongue in cheek internally, but to me, edge is the new cloud. Like remember, when cloud first happened, and it was this great thing, and it delivered all this value in these big players emerged, that is going to happen in the edge. There's going to be some big winners at the edge. And in my opinion, Rockwell with its global installed base, its 100 years of OT expertise, and its investments in digital solutions and partnerships, whether it's PTC, Microsoft, ANSYS, we announced recently, that's the hopes of this company, and certainly, in my opinion, and why we find ourselves so deeply engage with customers globally on our digital transformation initiatives.
Erik: So let's clarify the where the edge sits here, because I could see kind of three things being defined as the edge. One could be the sensors or the embedded computer in a particular device. And the second could be a gateway, something that has a bit more computing power and aggregates data from multiple proximate computers or devices. And then the third could be a server that's in a factory somewhere, and acting as a higher power localized data center. When we're talking about edge, are we talking about all of these and just deciding what work needs to be done, where based on how quickly that data needs to be processed, and so forth? Or are we really looking at one or two of these areas as where the innovation and the value is being created here?
Keith: It's all those things that depends on what industry you're in. If you make autonomous vehicles, the edge is the car. That's a data center, and you can't afford to go back to the cloud to figure out if that's a stop sign or a person. That's going to be the edge. That's going to be where data is processed. That's going to be where real time decisions are made. If you're a large automotive manufacturer, and you have data centers in your factories, that's going to be your edge, and lots of processing power and lots of ability to do two different things. If you're in fleet, your edge is the airplane, or the train or the mining truck. Part of the complexity of building solutions for this space is that the edge is very diverse by industry, and each of them have, to some extent, different requirements for the level of analytics and machine learning if they want to do, potentially, let's say an edge in the cloud.
Erik: Well, why don't we look at a couple of case studies, then we can maybe look from an end-to-end perspective on what this looks like in a real deployment? I shared with a couple previously. So one was a supplier to pharma and biotech, and specialty ingredients, and you were deploying digital factory solution. Can you give us an overview on what the objectives were here and then what that deployment look like?
Keith: At the highest level in the area of pharmaceutical, which is one of Rockwell’s strongest industries, if you will, there's a lot of things that go into the production of drugs. There's recipes. There's inventory. There's ingredients. There's everything right down to the capsule itself. And how do you operate a scalable production management environment tying all of those things together, especially in a world where we're moving to an environment where inventory management is really important, on-demand production is really important, and especially where you have in some areas have very high cost of goods?
And how do you keep these systems running, these manufacturing systems running at the highest efficiency for your production goals in line with your state of inventory, your cost of goods, and your customer demand, and minimizing things like inventory, maximizing things like yield having to do all that in a single system, and produce analytical insights for a number of different end users, whether it's the business unit owners, the people on the plant floor, the people who are responsible for production and distribution is the ultimate scenario. And that in this case, is what we're working towards with one of the world's leading pharmaceutical organizations.
Erik: . And so these are things where the status quo is that a company might have 10 or more different software suites that would be performing specific functions and be kind of operating in silos for a particular group of stakeholders? And you're talking about now being able to build a platform that integrates these functions and provides functionality for different types of stakeholders at different levels. How do you interface with the existing software that's on the ground? Is it a replacement scenario, or is it an integration scenario?
Keith: Most often, it's an integration scenario. These have been environments we've been working in and verticals we've been working on for some time. And so in many cases, there are some common partners from a software perspective that we've worked with, it most often becomes an [inaudible 34:26] into the existing operations. And as you said unifying data silos into a common view and then ultimately moving to a very detailed, so you need to keep track of a product through the entire production cycle with great visibility and great detail.
Erik: This maybe brings us to a question of the relationship between the platform in the past and the SaaS. So, I guess for me, it's a little bit unclear, we could be moving towards an environment where past solutions start to have more applications built into them from the software developer, and start to look a little bit more like a SaaS in terms of having really specific vertical applications integrated in. Or it could be more of a marketplace where the past is facilitating the processing of data and then you have a lot of different vertical applications that are provided by a lot of different vendors, and are somehow tapping into this. I guess, both for your offering, but also just across the market, do you see a clear trend towards one or the other of these situations?
Keith: Typically, early in a market, you're going to see more platforms as a service, and you're going to see more integrated solutions in, let's call it one system. Ultimately, you'll see more of a specialization, and an app store for the industrial IoT, and some of these different areas. Our partner, PTC has as a similar approach to that. In some use cases, I think that'll become more common as this market matures a little bit. I think, right now, customers are looking for fewer vendors to provide their solutions to get this at scale. And as this market matures, we may see more evolution towards the App Store models you mentioned.
Erik: And then there's one other case study here, which is a European dairy and here we're looking at a connected enterprise. So maybe we can go through a bit, how does this differ from the previous farmer case that we discussed?
Keith: Well, I think, this is an area where in this kind of dairy plant use case. But I think this is an example of something that we've worked on with Microsoft, with the focus largely on yield optimization. And I think you'll hear about this more and more in food and beverage. And I think, as you think about the need in the food and beverage market for things like mass customization, and minimizing the loss between production and retail, those are areas where there's a lot of analytics around what can be moved. And I think this focus on operational intelligence through the entire supply chain, and through the retail market to maximize the quality and the finished goods production, and the quantity of that, that'll become a KPI for anybody in this space. So for use cases that include both edge and cloud elements and that will be more and more quickly adopted by folks similar to this particular dairy producer that you're referencing.
Erik: I was just talking with the World Economic Forum last week, and they have an initiative, which is looking at data sharing. In this situation, you have the manufacturer, you have maybe a logistics provider, you have a warehouse operator, you have a retailer, who all have some data that's valuable to other people up and down the supply chain. And then the question is how do you actually share that data so that you can really get end-to-end visibility in a way that people are comfortable with that doesn't sacrifice IP, and so forth, and that maybe means controlling some metadata when you do that?
How do you view this situation right now? Do you think we have solutions? I mean, it's not only technical solutions. It's also contractual solutions that enable effective sharing of data. Do you think we've figured this out already so that it can be implemented more or less as a company would imagine in a best case? Or are there still a lot of either technical or contractual elements that still need to be worked out before we're really in this environment where we can share data between parties effectively?
Keith: I think that today, most customers who are doing these digital transformation initiatives view it as a competitive differentiator in how they're running their business. In this case, things like mass balance or overfill analysis or recipe management, that's a way to get to market faster, better, cheaper than your competitors. I think there will be a model in the future where we establish some form of trusted information broker. I think that's in the future. But I think it's a part of a maturation process of this industry that we're probably still little ways away from.
Erik: Keith, I think those are the main points I wanted to cover. Is there anything else that we missed here that you think is critical to cover today?
Keith: Erik, I'm exhausted with all those great questions. But I can't think of anything else. It's always great to talk to somebody who's neck deep in the latest in this digital transformation industry. I think you'll hear a lot more about Rockwell and Innovation Suite. And I look forward to following your work in this space. I look forward to next time we can chat.
Erik: Great, me as well. Well, Keith, thanks so much for joining and have a great rest of your day.
Keith: Thank you, Erik. Bye, bye.
Erik: Thanks for tuning in to another edition of the industrial IoT spotlight podcast. Don't forget to follow us on Twitter at IoTONEHQ and to check out our database of case studies on IoTone.com. If you have unique insight or a project deployment story to share, we'd love to feature you on a future edition. Write us at erik.walenza@IoTone.com.