Artwork for podcast The Industrial Talk Podcast with Scott MacKenzie
Ephraim Machtinger with Fortive
24th February 2022 • The Industrial Talk Podcast with Scott MacKenzie • The Industrial Talk Podcast with Scott MacKenzie
00:00:00 00:20:28

Share Episode


On this week's Industrial Talk we're talking to Ephraim Machtinger, Digital Product Leader at Fortive about "The Positive Impact to Industry through AI and Data Analytics".  Get the answers to your "AI" questions along with Ephraim's unique insight on the “How” on this Industrial Talk interview! Finally, get your exclusive free access to the Industrial Academy and a series on “Why You Need To Podcast” for Greater Success in 2022. All links designed for keeping you current in this rapidly changing Industrial Market. Learn! Grow! Enjoy!


Personal LinkedIn: Company LinkedIn: Company Website:


Accruent: Fluke Reliability:




NEOM Hitachi Vantara: Industrial Marketing Solutions: Industrial Academy: Industrial Dojo: We the 15:


LifterLMS: Get One Month Free for $1 – Active Campaign: Active Campaign Link Social Jukebox:

Industrial Academy (One Month Free Access And One Free License For Future Industrial Leader):

Business Beatitude the Book

Do you desire a more joy-filled, deeply-enduring sense of accomplishment and success? Live your business the way you want to live with the BUSINESS BEATITUDES...The Bridge connecting sacrifice to success. YOU NEED THE BUSINESS BEATITUDES!


Reserve My Copy and My 25% Discount


SUMMARY KEYWORDS ai, data analytics, industrial, people, fortive, reliability, bigdata, machine, problem, business, 29th annual, Fluke, Accruent, fleet, solve, headwind, conversation, called, build, manage, talk 00:00 On this episode of industrial talk, we are broadcasting live from the 29th annual SMRP conference in St. Louis. fantastic venue fantastic event. And we are brought to you by Accruent and Fluke Reliability. Let's get cracking with the interview. 00:21 Welcome to the industrial talk podcast with Scott Mackenzie. Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots, and let's go all right. 00:39 You're welcome. Welcome. And oh, yeah, welcome to industrial talk. We are broadcasting live from the 29th annual SMRP conference here in St. Louis, Missouri. Absolutely incredible venue with wonderful people. And I mean, when you start talking about the leaders in reliability and asset management and in maintenance, they are here from all around the world. It is a wonderful event. I also want to say thank you to Accruent and Fluke Reliability for sponsoring industrial talk. So I, I zip out here and I've got the the website and this is a statement that I just love with Fluke Reliability, provide maintenance and reliability professionals the data to do the job asset by asset wherever they are. Yes, yes. And yes. And then of course, with Accruent, we're talking about gain the insights to transform how your organization manages its physical resources. That is what today's game is all about. That is what we do here. And that's what SMRP And that's what Fluke and Accruent is all Accruent. A current effort a crew and crew and see. See, thank you. There's another one right there. That's a crew and he because he's with Fordham he knows how to sound. Yeah, you better not not me, you know, a crew it anyway. Hey, thank you for joining. Yeah, no problem. It's good. It's it's it's good to have a conversation with somebody from Fordham, which is the parent company, which is the mothership of both a current Fluke reliability to as well you go, Hey, we're going to be talking AI here digital transformation. But before we get into that conversation, because that's an important conversation. Let's talk a little bit about who you are a little background. It was a little 411 on who you're 02:27 sure so I'm originally from Canada. I grew up in Toronto. Oh, 02:33 just dropped. All my my parents grew up in Hamilton gas. And all my family lives in Burlington all of that area. 02:41 Wow. There we go. That's amazing. I don't need to say anymore. No, no. 02:45 Did you go Sudbury? Yeah, no, no, 02:48 I know of it. But I see the girl is probably the this guy. You 02:52 got to see the big nickel 02:56 but yeah, I I got into mechanical engineering as an undergrad, Major. Where'd you go to school, University of Toronto. Okay. And not McMaster, not McMaster, not anywhere else. And it was just because they their internship program was one year of continuous interning in one company, which was exciting rather than doing four months coops was pretty good. But I got into engineering actually, because I wanted to design rollercoasters, that wasn't that's my passion as a childhood, you see it out to build a huge automated roller coasters in my basement, still go to parks all over the country. That was the plan. And things kind of changed a little bit after I after I got into university. And, you know, I really got excited about industrial operations, just really designing big complicated systems, which really goes there's are that are interdisciplinary, require electrical, mechanical, engineering, etc. And, you know, I first had an interview with a oil and gas company after I graduated, which is called Suncor Energy. Biggest one in Canada. Yeah. So familiar with it, and went to go work in in maintenance planning and scheduling out in Fort McMurray for the huge cat 797. Yeah. And commodity 930 trucks example of very large, huge machines. And 04:16 that's that was that oil sands? 04:18 That was oil sands. Yeah, yeah. So we were, you know, that all supports the surface mining operation. And I did that for about a year and a half. And at that time, I was really excited to at this point, go back to GE where had done that one year internship at U of T. And I only wanted to do this rotational Leadership Program, which is called Oh MLP for operations management. And that was the first time in 2014, they brought it to Canada from the US. And I was part of the initial group with some friends of mine that were also interns and so I left and joined that program. And you know, that was a whirlwind of six months rotations in the oil and gas business of GE in the transportation business. And that was doing everything from lean manufacturing supply chain sourcing operations, supervision, and all that. And then I finished and stayed with GE transportation. And that's kind of where I started moving into the analytics side, because they have a very big program around. Analytics coming from fleets of locomotives, and they sell that as a service. By 05:19 the way, if you ever get put this on your bucket list, just look at a locomotive. There's some serious stuff. There's various serious tech 05:28 100 Is that a lot of it's pretty old. But 05:33 I remember I was able to get a tour of an engine. 05:38 Are you kidding me? Oh, yeah. Yeah, try standing next to one when it's running. It's just like, like, it's really It's powerful. 05:45 It's powerful. 05:48 But, but yeah, so I was I was working actually, out of our major customer CP rails office, I was the only GE person in their office managing that diagnostic program for their fleet of 600 plus locomotives. So we're really trying to crunch all that data and figure out what programs can we run for them to improve reliability the fleet and solve systemic issues, as well as identify them before, you know, before some catastrophic issue affecting the fleet. So I was there for about nine months. And, you know, I kind of fell a little bit of an island being just kind of myself and GE and the customer. I was like, where do we go from here. And I started, you know, looking for grad school programs and came across one at MIT, which was called leaders for global operations. And that basically fuse two degrees with which is an MBA and masters in engineering and obviously, pick your discipline. And it was right away, I was like, that's the place I gotta go. This is exactly me. I love operations. I don't want to leave that. But I want to take it to the next level, and explore emerging technologies in the field. And just, you know, be around like minded folks like yourself. 06:55 Yeah, I think that that whole conversation about emergent technologies that that digital transformation, journey, whatever that looks like that industry, for Dotto, it's just ripe with interesting solutions and conversations now, I was still in your, in your resume of who you are. The fact that we're going to talk about AI, where did you get that data analytics, that AI skill set? 07:23 Yeah, actually started in grad school. When I got to MIT, I had no AI background before. I mean, you know, strong in data analysis, but the field really was emerging at that time, and becoming popularized. And by the time I got to grad school, it was full blown. You know, there's the leading edge of research. So you can get exposure, you know, to the fundamentals, the math behind it all the way up to applications to all the startups that are leveraging AI in different ways. And it was just, you know, from nothing to complete immersion 07:56 over that two year period is a i, the way I look at it today versus the way it appears to be going or headed or, or the thought leaders are taking it isn't AI sort of this data, data analytic machine that creates parameters today, where if it's outside here, these are good parameters, anything outside of those parameters, bad got to do something. There will be a point where there is this sort of level of machine learning, but is is AI sort of at that level right now, where are we at? 08:29 Yeah, I mean, that's, that's what you described as pretty much how AI is applied in most situations where you need to learn a set of parameters that ultimately characterize the problem that you're trying to solve those parameters may not make intuitive sense. And that's part of the explain ability challenge. But ultimately, they can be used to generalize a problem. And then they can make predictions based on the data that you've been presented. So ultimately, it's really just a much more complex optimization problem. I always like to explain it as you have kind of a linear regression as the simplest kind of optimization problem that most people know fitting a line to a bunch of points and then making a prediction. And AI really does that in a nonlinear space. That's very hard to visualize. 09:17 Is that the headwind? Because when we start talking about AI, there's a number of questions that come up, one, it's going to take my job, I don't think that I think it's going to make your job better. I think it's going to address maybe some more mundane type of tasks that are currently being done. I think it's a it's a it's a positive, but there still is headwind of organizations and companies listening to thought and for lack of a better term hype like it's going to solve problems is going to do everything it's got to be everything versus what what it can do what is meaningful what what do you see it what's what's the main headwind? 09:58 Well, you know, think that like many kinds of technological, big technological changes in history, it's, you know, that hasn't kind of materialized right? When you think about the interstate highways being built in the US, a lot of people who were, you know, drunk driving carriages or cars for other people, and made a living off of that on doing short distances, thought they were out of a job. And after the highways were built, you know, all these other businesses started springing up in places that you couldn't access before. And that stimulated a new wave of growth granted some pain in between. And the same thing will happen with AI, but it most certainly won't replace people, they have to be augmented by people who can do higher level cognitive tasks, that AI today is just not built to solve. AI is very good at solving prescriptive, well defined tasks. But ultimately, there's still a level of interpretation and other processes around that, where the AI is used and deployed that have to be coordinated and worked on by humans. 10:56 The the AI in general has, is already in commercial use, in whatever capacity and and what comes to mind is how I'll say it, Facebook knows that I like watches, or whatever it might be. And then all of a sudden, lo and behold, I get watches on my stream, whatever it is, that is from my perspective, a function of AI saying, I see that guy's interest or, and I'm going to give more give and make it easier for me to see it. Yeah. In the industrial space, do you find where do we in this continuum? Where do we where do we stand from an industrial perspective? 11:42 Well, the I think what the industrial space, it's a, you're looking at different class of AI problems, right, and a lot of what you're describing, it's recommendation problems, and it's very well suited to e commerce, when you're trying to buy things. In the industrial space, it's a lot of time series data that you're dealing with. And so rather than make recommendations about what to purchase, or what to do, you're you're trying to diagnose or detect that something has occurred or is about to occur. And it's, it's a lot more, I would say, technically challenging problem, especially when you're dealing with the kinds of signals that you're gonna get in an industrial context, much more complex than, say, an E commerce 12:23 scene, it is that complexity that always sort of overwhelms me, because you're absolutely right. If I have an asset that's out in the field, I have a device on that asset, IoT, IoT, whatever, whatever the device is, it's collecting data, it's sending it to some sort of edge solution into the cloud. And that's where we're trained in and hopefully, be able to, maybe have some sort of computing capability at the edge to say, that's, that's dummy data, or whatever that data we don't care about that data we don't care about, we want to let this data go through. And it's, it's an interesting, it's just for that pump. There's another pump, yeah, there's another motor, and each one has specific data that is meaningful for it. And it even gets down to probably to have a pump a is same as pump B, but the performance is just different. Exactly. And you're gonna have to constantly manage that to come up with meaningful tactical, right provision, 13:24 and you have decisions to make, right when when it comes down to doing the number crunching and analysis. You know, some techniques tried to generalize over similar machines, but operating in different environments, and figuring out how to, you know, provide Accruent diagnoses, irrespective of the environment they operate in, which is very challenging, but very useful if you can do that. And then others treat as you said, the machine is all unique, which they are they have their own data streams, even if they're identical machines, and they build systems around each machine. But now you have to be able to do something that scales very quickly across a number of machines, because you don't have the time and resources to do have a complex conditioning and training process for every single machine. Yeah, 14:10 it's it. But but but it has benefits. Right? Can you explain a little bit about the benefits of where you're going with AI and that sort of digital transformation journey? What Why is it so important? 14:27 I think broadly speaking, what you're seeing in the marketplace is a lot of concentration of companies, right? A lot of the big tech giants are ones that come to mind. Yep. And simply the way that that's been abled is through digital business models. Digital Business models have effectively reduced the incremental cost of delivering value to a customer to zero. So if you're Amazon, what's the cost of serving a new customer? Zero. I don't have to build a product. I don't have to ship anything. Nothing, they they just, you know, provide some additional data flow back to in front of my, my site. And that's about it. And so the the, that lack of any cost, but also the scalability, that digital networks offer is amazing. And it's what's allowed this concentration to happen. And so, you know, when you're saying why digital transformation is so important, what you're trying to do is unlock that scalability that is very, very hard to achieve.