Gary Mintchell from The Manufacturing Connection joins Don Pearson to give us his insights on the trends he’s seeing with as-a-service software solutions and the specific value they bring from both the IT and OT perspective. Gary shares the importance of the cloud in relation to edge computing in the industrial space and designing your architecture to get the best of both worlds. They discuss how to utilize OT and IT together to solve common problems, debate whether MQTT and OPC UA are in competition or complementary, explore the areas which companies should focus on in their journey to digital transformation, and make predictions about AI and machine learning. Gary also shares a memory about meeting our founder Steve Hechtman when Steve was still an integrator.
"I never look at a technology and say, 'Let me fit it into the company.' It's more like, 'I've got a problem to solve, now where can I go find the technologies and the products that are going to help me?'"
Gary is a passionate communicator and adviser on technology trends focusing on manufacturing, marketing, leadership, and personal development. In his business career, Gary has held leadership positions in almost every area of manufacturing including IT, product development, and marketing. He has acted as an angel investor, launched new products, increased sales, managed projects and programs, helped companies and individuals succeed, and was among the first in his market to deeply engage in social media. He was one of the co-founders and founding editor of Automation World magazine. In addition to being a certified yoga and fitness instructor, Gary is an accomplished soccer referee, having refereed at the college and adult level as well as working three state high school finals.
Don: Well, it certainly is my great pleasure to welcome you, Gary Mintchell, to the podcast today. I think I'd like to start off with a little introduction that you give of yourself. We've known each other for years, we'll talk a little bit about a little background and then get into some topics. So, Gary Mintchell, welcome to the podcast, and tell us a little bit about you and what you're doing.
Gary: Hi Don, it's great to be here. It's actually kind of strange me being on the other side of the interviewing thing. I've been podcasting since 2007, so I often tell people I actually worked before I started this gig that I'm on now. I worked in manufacturing and product development, manufacturing engineering, quality assurance and stuff for about 25 years, and went into publishing and became an editor, and I think most people knew me as the founding editor-in-chief of Automation World Magazine, and I left to go out on my own... Oh man, it's eight years ago now. And so, I do The Manufacturing Connection, I'm an independent blogger and analyst type of a person, and mostly I just like to promote the industry and good technology and good people of which I know a bunch, and you're one of them. So, you and Steve and the crew out there, I just enjoy working with you guys.
Don: Gary, it's great to talk to you. You sort of shoved me down memory lane when you started talking about yourself too, because I think it was... I don't know, it's 15 years ago now, at an ISA show in Houston, I think, may have been the first time that you and I met. That's a while ago now. It seems like... A little bit it seems like yesterday and it all seems like a long time ago, but then I was also thinking when you said that about you being on the other side of the microphone, if you will, right now. I think you came to our offices, Inductive Automation offices, the old, old offices when Steve was still in the integration business, and you did a podcast interview with him almost 15 years ago, is that... Is my recollection correct or is that... Am I totally off base on that?
Gary: Oh no, we... That was interesting. I had met Steve, the founder, a couple of years before that at an ISA show, and didn't know how to take him. He just said, "Got this new company and we're IT from the ground up, just totally rewrote everything," and I meet CEOs like that all the time, it's kinda like, "Okay." But anyway, I listened to a while, we talked and so on, it was very interesting what he was doing at the time and... Yeah, so we made a trip out to Sacramento and I remember those were the early days. I laid a little digital recorder on the conference room table, thought I was going to interview Steve, but Steve is a quiet kinda guy, and he just turned it over to... Who was there? Travis was there, and...
Don: Travis, and maybe Colby, I'm not sure, it may have been...
Gary: Probably Colby, I think. Anyway, yeah, you could find it at the automation.libsyn.com, and somewhere back about that time, it's still on the web. But we had a great conversation and just those guys talking geek, and Java and different things and what was going on. Yeah, so it goes back a long way. I followed you guys a long time and you've done some really interesting stuff out there. I can't believe it's been that long, but it's been a while.
Don: And I was just thinking when you said that, in those days when we were at the ISA shows and we were ground up Java, the Ignition platform was just in early development stages, we were still working with legacy products. I remember talking at shows and Steve would be talking to other CEOs and they'd go, "Why are you trying to create a SCADA product? I mean, that is a mature market, there's no room. Nothing's gonna happen there. You're not gonna make it. You're selling things too cheap, you're gonna go out of business," all the various things that admonitions from other folks in the space and now I look back over 15 years, and the world of digital transformation and the world of software aiding the world, as Mike Milinkovich from the Eclipse Foundation has been known to say, the world has changed. The truth is there's a lot of market out there, and a lot of stuff going on, and it's great to have a chance to chat with you about some of it because you've been in this world for a while and you're still in touch with trends and things going on.
Don: I think... I think I'd like to start maybe with just taking your experience, what are kind of the trends that you see when you look at anything as a service? Software-as-a-service, platform-as-a-service, whatever it happens to be, that whole world, you have some thoughts, opinions in that area that we can, sort of, build on here?
Gary: Well, we can start off on that. I did some research, actually to broaden the as-a-service idea for another panel discussion I was on recently. And I found out that there is no definition. There's a famous movie from years and years ago called “Christmas Vacation” and Clark Griswold, the famous lighting guy, explains to his little niece, you know, Christmas means something different to everyone. And I'm thinking, this whole as-a-service thing sort of does, depending on who you read, but I think there are a lot of different ways we can go and listening to vendors of various kinds, I mean, you guys, big vendors and small vendors, and some customers, it goes a couple of different ways. And the interesting thing is, you know, it almost kinda got really popular with Salesforce and customer and CRM using software-as-a-service, meaning you didn't download and use the software like you used to with Microsoft Word, for example, or WordPerfect, if you're an old guy like me.
Gary: You actually had it on your computer and you thought you owned it and all that, and this was a whole different way of doing it. So, part of what I see and part of the trends are using a product where you don't have to load it all on all your different computers. I think it's called buying seats, where you have to buy X-number of copies in the old days, you bought floppy disks or whatever, and then you would stick it on an individual computer and all sorta stuff. So, one of the things I see is kind of this idea of product almost more like a... An operation expenditure, OPEX, rather than capital expenditure, because I don't have to buy some big, big expenditure and then download it myself. And so that's one of the things I see. And there's a lot of growth in that. There are companies pivoting. You guys showed the way on a lot of this stuff. And then a lot of your competitors have pivoted, as they like to say in Silicon Valley, in that same way. There's another aspect maybe we can get into later or maybe we're out of time, I don't know, and I go back two decades, these things never occur instantaneously.
Gary: And a thing that I first saw, a nice little kinda control board with a lot of memory, and I/O in a Nokia cellphone module on it, and it was called machine-to-machine, M2M. And the idea there was, I could put one of these boards in a machine, bypass the control system, so I don't have to get the control engineer's blessing to change a program or anything like that, or getting into the data table, and send some data up to the IT department. And the idea then becoming, one, IT can get into it. But the big sales pitch was going to be an OEM. A machine OEM, skid OEM or somebody like that could embed one of those things in a machine and then sell services. So that would be another as-a-service kind of idea. Two decades ago, the technology wasn't there. I used to say, "Your cell phone's dropping calls all the time. Can you hear me yet?" That sort of stuff. But now, the technology is there. So I see a second trend kind of going... Revisiting that whole idea of, can I sell some other kinda services? Quality-as-a-service or predictive-maintenance-as-a-service or something. So I see a couple of different things going. It's very interesting right now.
Don: Yeah, it really is. Do you think it has a different... If you take an IT top-down viewpoint of that or an OT ground-up viewpoint on it from the field or the plant floor, do you think the viewpoint of what the value is and what the interest is is different if you're looking at it from an IT or an OT viewpoint?
Gary: Well, yes, it is, because IT needs to feed the IT... Their systems. ERP and sales tracking and things like that. And they need... The promise, more than two decades ago, was, "We'll get all this manufacturing information, we'll be able to help executives manage their companies better and manage their plants better," and so on with all this data, and then it didn't happen because it was too hard to do. So a lot of this technology, because people like you broke away into what's now called cloud and using the internet and that sort of stuff, made it easier to feed, what I would... What I call feed the beast. And so that solves an IT problem without screwing up the operations people, which is really nice, because don't we always like to talk about the fighting between IT and OT? And I always thought, "Well, I worked in both, and there's no inherent reason to fight." So the OT people, their problem is, it is kinda growing. It's more than just keeping a machine running now. It's keeping the whole line running, or maybe I can manage a whole plant better. So now I can start talking... Now I can do more and make better decisions and do better things. So it works both ways. And I see... I'm sure you see more than I do of... Because I've pinned Ignition back when it was live, and talked to people on both sides, and now they talk together and they're solving common problems.
Don: Yeah, I think you're right. I think the OT, IT folks have come together just a lot better over the last few years, and they're beginning to realize more that we have one enterprise. Another angle on it, though, that I might get your comment on is, in the industrial and manufacturing space with those kinds of organizations, and you look at the OT space, there is a difference between operational data, real-time data working... The dangers, if you will, of some of what goes on in the manufacturing space and having things go to the cloud, not historical data, not for that reason is, do you see... Do you think... You think this sort of move to moving things to the cloud, software offerings that aren't locally installed, as you're talking about, do you think it kind of is a direction that we should embrace in the OT space of the plant floor, and if so, why, and if not, why?
Gary: It's kind of interesting. I had a customer back in the '90s who said, "I will never allow a wire to come from a PLC to anything, other than I/O," of course. And...
Don: That's the starting point.
Gary: And I think about that all the time, as we do all this connectivity. But we had to solve a lot of problems along the way. OT, when you're working where a machine can injure somebody or a machine can break and put production down, I've been on the line with the plant manager is saying, "You know how many hundreds of thousands of dollars I lose a minute that this machine's down? Get this thing up and running." I've been there. So you've got that kinda consequence, and you don't want people probing around in your system. And so then people started talking about cloud, and the first thing the operations people thought was, "Oh, controlling the cloud. We'll never be able to do that, because of the time lags and things like that. People breaking in, security becomes a problem." And so how we solve all that problem is, security is a bit of a moving target, but security is an awful lot better than it used to be, because we think about it.
Gary: So that's one thing. And the other thing is, we don't need to move control to the cloud, necessarily. I'm sure somebody's dreaming of that. But if you've got the control as a black box, there are a lot of reasons to move data into the cloud. And the cloud simply is a compute platform somewhere connected to your stuff. And it could actually be in your own facility, it could be in a corporate quarters, it could be Amazon Web Services and Google and Microsoft. Everybody's got a web platform, one of those now.
Gary: That handles that part of it, and I was amazed years ago, running into engineers who were doing their own serves. They'd just go out and like us old guys, brought a PC in, they brought servers in, and heck with you PC guys. And, "Oh okay, this is interesting." What happens is, we adopt these things and we learn how to use them, and so, that becomes very interesting. So, there's a reason, historical databases, so I can do trend analysis, from trend analysis, I can do predictive analytics, and there becomes a ton of operations reasons to have all these data stored somewhere outside of the control system.
Don: I remember having a chat with Arlen Nipper from Cirrus Link Solutions when he was early on... They were talking cloud in their oil and gas industry, and he shares the example of just not that many years ago, talking to folks at different oil and gas companies, going, "You will never see us go to the cloud. It's not gonna happen. It's not a possibility. Never, not on our dead body." And yet, look at five years later. Everybody's doing that. So, I think it's moved strongly in that direction. I wanna shift just a little bit the attention to the world of cloud computing in relationship to edge computing. There's a lot of movement and moving things closer to the edge, and then there's the movement of things in the cloud. Can you give your view and elaborate a little bit on the importance of edge computing and cloud computing in the industrial space? And maybe a little... A comparison of your thoughts on what should be done where, from your perspective?
Gary: Interestingly enough, I named my website, and technically my business, The Manufacturing Connection, and the reason was... Well, two reasons. One, I could buy that domain name, and the other is, there was connection. I was more interested in connection, and the adjective was difficult to come up with. I've always been involved in this connectivity thing, which led me into the realm of IT as the only independent blogger or influencer analyst kind of person, not beholden to anybody. So, I'd go on to a lot of Dell conferences, a lot of Hewlett-Packard Enterprise conferences, it's interesting, as the manufacturing guy. Sitting with HPE, a whole 20 influencers and other technical people, it's like, "What's the edge? Let's talk about the edge?" And they'd look and say, "Gary, you are the edge."
Gary: Everything you are is the edge of the network.
Don: The entire manufacturing space is the edge to them.
Gary: Is the edge, yeah, because those guys are all the server guys and the storage guys, and the big iron guys. So, why do we go there? We just talked about cloud, and cloud is a server, it's a big compute platform somewhere, and it could be rented, you could own it, whatever. But we got another problem is, in manufacturing, we generate a lot of data, as in a lot of data. If you're gonna transmit that data, what does it take? Bandwidth. Bandwidth isn't always all that available or cheap, or whatever, so then we get into this idea of edge. And it started out with edge gateways. Can I connect a whole bunch of stuff to some kind of a box that's got some storage in it? And then, I can ship it off to the cloud. And then, they say, "Well, wait a minute. Why don't we... " Like in Hewlett-Packard Enterprise's case, they bought Silicon Graphics and Cray, and they've got compute power beyond imagine, almost. "Why don't we put that in a box?" And they've put them on offshore oil platforms, to talk about your oil and gas stuff, so that they can do a lot of compute there and storage, and just ship what needs to go to the cloud to reduce the bandwidth problem and utilize the best of the cloud, utilize the best of the edge.
Gary: People get confused, "What is edge? And can I replace cloud with edge?" Well, maybe. I'd look at it like, "Look at a system. What problem are you trying to solve? What technology, then, can help you solve your problem? And then, how do I do a system architecture that takes the best of all these worlds?" And that's looking at the hardware side, the edge and the cloud server, and then you look at the software side of, "How do I load and communicate and so on?"
Don: Yeah, when you look at that, do you see these, then, as competing technologies, separate solutions, or as I think you're saying, part of an integrated architectural development, to do the right thing at the right place with the right technology? How do you see that evolving?
Gary: I don't see so much competing right now. Sometimes you get evangelists, and Guy Kawasaki is the famous first evangelist for Apple Computers a long time ago, and now we've got 'em everywhere. But sometimes, they go a little overboard in promoting this or that cloud or edge or whatever. What I see is, almost everybody I talked to in the last two, three years realizes this hybrid approach where there's a reason and a purpose for compute at the edge, a reason and a purpose for compute in the cloud, and we just design our architecture to take advantage of the best of both worlds. And I see quite a bit. I've talked to a couple of automation vendors lately, and looking at things like the PLC as the edge compute, which blew my mind, 'cause I'm still an old-school guy. When I went to school on PLCs, it was Allen-Bradley's PLC-5. When you're old, you're old, but there's a lot of stuff there. I'm seeing a lot of people do this edge thing and process control people are buying their private labeling, or whatever it is, some sort of good compute platform in a box that they can put on a offshore platform or somewhere. I see less competition and more, like, "How can I do a system and how can I put it together?"
Don: That makes sense. You mentioned something about... Yeah, when you start looking at moving everything up and the volumes of data that exist in the manufacturing space, leads me into another topic I wanted to get you to give your opinions on is, MQTT is becoming, as a transport vehicle, very, very popular in the industrial space. Obviously, our strategic partner with Inductive Automation, Cirrus Link Solutions, its president is co-Inventor of MQTT, Arlen Nipper and we've worked with him for years, but there's some talk about OPC UA, MQTT, is one supposed to replace the other. Are they a competitor? Do they have a play? And it's just where do they play in a similar thing. Solving some of the bandwidth problem has to do with... Obviously a vehicle for doing that is the report by exception technology of MQTT, so you're not wasting bandwidth on traditional poll-response technologies. So when you look at MQTT and OPC UA, from your perspective, is it a competition, is it again complementary, you use one where it makes sense that it allows you to have a more robust architecture overall? What are your thoughts about those two approaches and evolutions in the manufacturing space?
Gary: That's a very interesting topic. And I have a weird part of my personality where I like to stir up things occasionally.
Don: I've noticed that about you, Gary. So that's why I asked the question. Go ahead and stir it up.
Gary: Yeah. Well, I don't wanna get in too much trouble, but it's interesting. So I've listened to Arlen. He's one of those patient evangelist kind of guys I just mentioned. And I think he's got that same quirk in personality, because he'll kind of smile at me after he makes a bold statement. But I have a lot of friends in the OPC arena, of course. So I would write to the OPC people and talk about MQTT and Sparkplug and those technologies and try to get them all riled up and then you get a response. So there's an evangelist of both sides. So some of the... There are OPC UA evangelist purists or something, and they talk about OPC UA to the cloud, and they can replace MES or it can... I don't know, solve world hunger, I don't know, a lot of stuff. There's always somebody like that. The weird thing is that OPC UA really is an information model. And it's really useful. It's one of those weird things that almost everybody uses and almost everybody complains about. I love those things, it's like... But there's a ton of uses for it, but maybe not every use. But OPC UA needs a transport mechanism.
Gary: And when they first did OPC, they did it over something I think is called AMQP, which is just another transport protocol. But it also works over MQTT. You can do that... So in that case, there's no competition, it's just... It's a couple of different things. So if we look at that part of it, then that's one part. But then the MQTT people, you know Arlen and his group came up with another... With a model not nearly as robust on purpose as OPC called Sparkplug, and technically you can make a point that Sparkplug and OPC UA compete. But in another one of your partners I see at Ignition Community Conference all the time is my old friend Benson Hougland from Opto 22, and he's always done IT/OT but he's always taken a different take than anybody else. And so he and Arlen and you guys kind of put together a simpler, easier to use, I guess is what I would say and that's not being fair, but it's good for this podcast. Because you don't need all this other stuff, but you need to solve a particular problem. And so MQTT Sparkplug to Ignition and your various things works wonders for a lot of the applications. So if I go back to what's the problem you're trying to solve, if it's a lighter-weight, kind of less information-intensive thing? Perfect. Maybe you need all the information modeling of OPC UA. Well, gee, we have a solution there where I just read about your new 8.1.something?
Gary: Well, what I was reading about it, is it set it off a lot about OPC UA supported, it's not like you guys support. What do your customer need? Okay, we can help there.
Don: If you go back earlier, even prior to any work with Arlen and the team at Cirrus Link and MQTT modules, we based the company on OPC UA before it was a standard totally out, 'cause Steve was well aware as our founder and CEO was calling, Colby that, "We need to have data, we needed to have connectivity, we needed to have all the things that OPC UA was offering." So clearly OPC UA has a significant role to play, but you're right, you need transport too. So I guess that was my basic question, do you see these things as competing, or what should the relationship be between OPC UA and MQTT in the manufacturing space from your view?
Gary: I think on one hand, they're complementary and I think on the other hand, there's a level of competition that I view as good because it keeps everybody on their toes. It can keep the OPC UA people stretching and looking at what's going on down here and what's needed. I used to be in product development. So I think of this, I'm always looking at the competition, and say, "Okay, competition is good. What can I learn? What can I do? Does it make me better?" And the same thing with the whole MQTT Sparkplug thing. So they're both gonna exist, and they both fulfill purposes. And I think they are somewhat competitive. And I think that's good. I think it's good. Full stop. Yeah.
Don: Well, then let's shift a little bit. We talk a little bit about as-a-service and the world of cloud computing and the evolution in that direction, the view of edge and cloud back and forth, and then transport in the OPC UA/MQTT world. When you're talking about an organization that's looking for an industrial solution, as a broad stroke, what should be the main focus they have, the technology, the product, what the product brings to the business, when they're trying to move forward in their own digital transformation journey, however their IoT solutions, whatever. What should they be looking at as a broad stroke when they're trying to put together their strategies and move forward?
Gary: Oh, gee, that's just a nice narrow question, Don.
Don: Could you do that in two minutes, Gary?
Gary: Okay, we'll go for it. Tick, tick, tick, tick. No, actually, whenever I'm advising people, people come to me and they'll ask, "Well, I think I should do Internet of Things. IoT." "Oh good, why?" You know, it's kind of like all that. And so I actually take a step back and say, "What's the business problem you're trying to solve? What do you need to do? Do you need data from this? Do you need to compile data? Are you trying to do... Improve your quality, or predictive maintenance?" I've got some kind of a problem I'm solving. And then I go out, and I look, and I say, "Okay... " So I never look at a technology and say, "Let me fit it into the company." It's more like, "I've got a problem to solve, now where can I go find the technologies and the products that are going to help me?"
Gary: So that's step one. So then I start looking. And you're gonna need a partner, and you need somebody with staying power, and you need somebody... And you need to look at products that are built such that they can expand, they are scalable, that it's a company that brings out new versions, and even better, started this whole conversation with as-a-service, and one of the as-a-service things is if I let you host the thing on the cloud and I buy just the services, I'm always gonna be up-to-date because the vendor is going to be keeping it up-to-date, and all those benefits that come from it. So if I'm gonna look on the software side of things, then I'm going to look for that. If I'm looking on the hardware sort of side of things, I wanna buy things that divorce hardware from software, for example, so that I can upgrade one without screwing up the other and all that. That's a big thing in process control right now. And so I can look ahead into that sort of thing. It's a hardware/software... Yeah, I think those are the kind of things I'd want, somebody I can work with that's gonna scale.
Don: A partner with some staying power, and also staying up-to-date, you're right, not having to worry about keeping up-to-date yourself because the as-a-service side of it gives you some confidence that it's kept up-to-date. I wanna give us one final shot for you as we sort of wrap this up a little bit. This is another broad question for you, but you always have opinions. Ever since I've met you, you have opinions about where things are, and where they're going, and where they should go. So when you look from your perspective out into the future, what are some of the exciting things that you're interested in, that you see coming, that get you excited about where things are going in this overall world of manufacturing connections, to your name, and the industrial space broadly?
Gary: You know, I was just beginning to wonder if things were getting in a situation of stasis, that where's innovation gonna come from? Little tweaks as we upgrade and make our products better. This whole idea of artificial intelligence, it really gets overhyped, and it's really crazy, and it's not like popular press might have it, but there's a lot of AI built into stuff that's going to be very interesting and useful. And you may not even know it's AI, or maybe not even care, but there are some companies springing up, doing a lot of stuff in there. I think this as-a-service... I approached it at the beginning. Right now people are talking about product-as-a-service in the sense of, "Let me put my software on the cloud and then you can buy instances of it," or I don't know what the proper term is. But I think in the future we're gonna see companies, and maybe even you guys branch out. I don't know or some other partnership where you can start looking at quality-as-a-service, or predictive-analytics-as-a-service, or something where you sell something, or you sell like through the machine vendors, or the skid manufacturers and that sort of thing, where they can really build this relationship of servicing my products and systems as an ongoing sales revenue thing, looking at the vendor side of it, how do I make more consistent revenue.
Gary: And then you look at the customer side of it, I can't hire people anymore, right? It is so hard to hire anything, from engineers, and technicians, and if I can rent them that works nicely too. There's a lot of stuff going there. I'm excited about that. I think there's a lot of cool... There's some new companies starting up with the edge, IoT, AI combinations of playing with stuff. I think we're gonna see some really interesting responses to what's happened over the last couple of years, and then take us into the future.
Don: It's a good time this time. An exciting time too. I certainly see the evolution in the demands on ourselves as a company as exciting, 'cause that demand also puts opportunity right there in front of us, if we can keep up with those opportunities and address the market. There's still a lot of room for innovation, I think, to your point, so. Well, listen, I appreciate having the time to sit and chat with you. I've been asking you a bunch of questions across a variety of topics. I don't know if you've got any parting words, or something that doesn't connect to a question you wanna say, and as we wrap up, but anything else? Anything final?
Gary: We're recording this before Ignition Community Conference. I'm sorry I'm not going to be there, but I'll be there in spirit, and virtually, and hopefully, in another year we can get out there and bring that whole community together. They're always a fun group.
Don: Gary, that would be great to see you. We're certainly optimistic. We have a fully virtual with ICC again this year, but we're very optimistic that as we look to 2022 we will be able to have a hybrid conference, and we can have you come out, and we can have some opportunity to interact again and have the virtual side of it, but also have an in-person side of it. So we look forward to seeing you again at that time. And I just wanna say thanks again. It's always good to have a chance to chat with you. I'm glad you're doing well, and it's really fantastic that you and your wife are located there where you can hang out with family, and see grandkids, and all the things that I know you and I both appreciate about the family life side. So thanks again, appreciate the conversation today.
Gary: You're welcome. Thanks a lot, Don, see ya.
Don: Have a great day. Bye-bye.