Terri Lewis is a senior executive at Planet Connected and a 32-year veteran of Caterpillar. She began her career in technical field support before embarking on several "mini careers," including aftermarket parts marketing, service operations, product marketing, heavy construction marketing, product development, M&A, electronic supply chain, and end of life support. Terri then spent six years in digital strategy where she learned how digitalization could completely transform a business, from customer activities to client interactions and overall company goals. Plant Services editor in chief Thomas Wilk and managing editor Anna Townshend recently spoke with Terri about how to successfully implement digital transformation initiatives.
PS: So Terri, it's been about a month and 1/2 since the Fluke event. You've done a little bit of global traveling since then. Hopefully we can touch on that, but let's go ahead and start with your time at Caterpillar. You told us a couple of really great stories about that time at the Fluke event, and the detail that stands out is that you framed your experience there as eight different career changes without changing company. And that was in large part due to technology changes. Could you talk to us a little bit about that?
TL: Yeah. First off, I’d like to compliment you for saying that I'm younger. Actually, I spent 32 years at CAT, not 20.
PS: Ah, my mistake, sorry Terri.
TL: No, that's alright. I just feel young! I had fabulous opportunities at Caterpillar. I started off in the technical field support when I hired in. I was an electrical engineer, a recent graduate, and at the time Caterpillar had 65 electrical engineers. It was not big. I moved into aftermarket parts marketing, service operations (you know, efficiency studies in terms of the repair areas or the dealerships), product marketing, heavy construction marketing, product development, M&A, electronic supply chain, and end of life support.
Then the last six years was in digital, in the strategy role set up by the executive office. All of those (I call them) career changes really helped me understand the impact that digital could do to transform our customers, what they do, our interactions with them, and transform what we were doing as a company.
PS: You mentioned specifically at the event that your focus was on eliminating downtime. That was your passion while at Caterpillar and it still kind of is, right?
TL: Yeah, absolutely. I think that the really interesting stuff is taking data and technology and knowledge, right, and my take away was making unplanned downtime a myth. And that seems a bit bold, but I think it's doable, right? It's bringing together the elements of people, process, and technology together, and leveraging that, starting with people – emphasize people, right, it's not just about tech – and I think most companies can get there.
PS: Does it take the same time for companies to get there or do you think that some companies are always looking for one of those magic three – people, process, technology – and there's always something that's missing?
TL: Yeah, that's a great question. One of the things that I see is a mistake over and over, including what we did at CAT, was focusing on the technology first and not understanding people and process, then technology, in that order.
I was excited at some of the other presentations by customers of Fluke. The ones that were successful focused on people, right: get their engagement, help them identify what the problems are. They are really the secret sauce of an organization of knowing what the issues are and then watch what they're doing from a process flow, and then figure out how to leverage technology to make it work. So yeah: people, process, tech, in that order, not the other way around.
PS: Klaus Blache at University of Tennessee Knoxville's Reliability and Maintainability Center is fond of saying that every success story of this kind of starts with maintenance and operations getting on the same page, and understanding what the problem is and what they want to do together and cooperating.
PS: Can you talk about one project at Caterpillar on the digital side that was a success story? Something that you could point to as saying, yeah, you know what, Caterpillar and our team recognize what digital could do for us and we got to win out of it?
TL: Yeah, let's see, examples. One of them that's out there that was really interesting was in Namibia, B2Gold, they were actually trying to improve their sustainability and renewable energy, so they put in a solar plant, solar field with gensets. We ended up connecting it, all of it. We had about 24,000 data channels coming back to us with weather data and energy off of the systems, the processing plant, to help them optimize their energy use in a remote location. It's out there on the Internet in terms of what we did, it's public information, but it really helped them in terms of their sustainability goals. Getting fuel into remote locations like mine sites is expensive, sSo it helped them reduce their costs and you know their operations. So that was one of the most difficult, but it was one of the coolest projects I worked on.
PS: That is really cool and we're living in a sustainable moment too, where that story is going to resonate with people. Mine stories are always interesting to me because once you make one bit of progress in one area, usually there's like a rollover positive effect on other areas. I've heard of mines that sensored up their fleet equipment and sent a lot of data via the IoT to the cloud to figure out how the equipment is operating. They were so successful at understanding variances in sensor data that they could tell if this was a vehicle issue or a pothole in the road issue, and they actually ended up being able to fix the roads more efficiently because they could tell when the anomaly was a pothole. It ended up not being just fewer repairs in the fleet itself but also greater production, because the roads were better out of the mine and they were able to move more ore more quickly.
TL: That's a good example. Going back to that B2Gold application, right? We had the data from our gensets – we were connecting thousands of our gensets, and those are big power, if you don't have power, if they're not connected to the grid, it can be life changing depending where they're at. We had our predictive reliability working on our gensets to help them prepare and order parts because they're remote, and didn't want to have inventory of everything out on the mine site. We helped plan ahead on the maintenance and repairs and then making sure that they always went on. We could do the remote Start / Stop of the generators and take that data, apply our analytics to it, and understand whether they were going to be ready to run when they needed to run.
PS: Sounds like some amazing work you got to do at Caterpillar. I'd like to shift the focus a little bit to your current work. I know you work a lot with small to medium sized businesses with their digital transformation initiatives. So how about you talk to us a little bit about common projects you see, or ones that are really interesting? What's going on out there?
TL: For everyone who's worked at CAT, we've garnered a lot of experience. I would say our digital transformation or IoT started in 1994 when our one of our vice presidents of aftermarket parts created a set of videos for executive leaders that gave a viewer a vision of what IoT could do for us and our customers and dealers. It's on YouTube and it's called Blast from the Past. And when people see that, they realize how innovative and forward thinking Caterpillar was. The perception is we were a boring sluggish industrial company, and it wasn't true.
So we really embarked on that and I think over that time frame we made collectively every mistake possible. Which is not a bad thing, I guess, as long as you learn from it. And one of the things I like to do is to work with small and medium sized businesses to help them, to avoid those mistakes that we made and learn from the experience, because a lot of them can't afford it. So that's kind of the gist of it, was to share that experience and knowledge to help others grow.
PS: Do you find that those small to medium sized businesses are really facing sort of the same challenges or the same problems that a big company like Caterpillar faces? Are they different?
TL: I think that they’re the same in a lot of ways. Maybe one of the differences is a large company can maybe spend a bit more money, right? Small companies have cash flow issues, a lot more than an established company, but one of the things is, technology is always a challenge. You know, the right vendor, the people, the process. I think one of the other things that is a challenge for everybody is commercial model where we've seen IoT deploy things. Personally, I think one of the toughest change issues is commercial models. How is a small company going to make money? Going forward is different than it has been in the past. Maybe it’s the right way to do it, but then you got to get customers to accept that.
One of the stories from CAT was we were trying to understand how our connected products were being deployed. We were looking at who is using the data, and we thought it was the small customers, then we thought it was the medium sized customers, or it was large customers and maybe it was landscaping, maybe it was construction. None of that rang true. What it came down to was the customers that were comfortable using cell phones were the ones that we're using the data right? They saw value in the data and they would actually pay for it. So I think that's the same thing for small and medium sized customers, is who's going to value it and what's the commercial model. It's probably going to defy historical logic sometimes.
PS: That's fascinating how the ubiquitous consumer technology made a big difference. But I hear what you're saying too, your framing the issue for me in a new way, which is that some companies simply have less margin for error when it comes to budget expenses. Some companies don't have 3-4-5-6 bolts to fire, they've got one or two, and so hopefully it's got to work out the first time or second time.
TL: Caterpillar, we had a global reach, right? So we had the market intelligence, we have our dealer organization, it's about 100,000 people at CAT and another 100,000 people at the dealership. We always work with them to garner that. We had that market information – boots on the ground understanding, listening and talking to people – that a lot of small companies don't necessarily have.
PS: Let me look forward to a different topic. I'd like to talk to you about AI. We both would, but let's start first, I want to congratulate you. You've just started a program in Smart Cities at the University of Central Florida. Congratulations!
TL: Ohh, thank you, thank you! I'll start in the fall. My kids are wondering, what mom doing going back to grad school? But they think it's pretty cool.
PS: That's really cool. Now, is this a Master’s program that you helped create with the UCF team, or is this an offering in smart cities that one department of the university offers?
TL: I was looking around for programs and the University of Central Florida has been one of, if not the leader in developing this program. It's part of their civil engineering department and it's smart cities. For five years I've been doing some fabulous research, taking a traditional civil engineering degree, matching it with policy, matching it with electrical engineering data and analytics, and bringing in some of the core programs and issues for smart cities. So I'm very excited about the program.
PS: That's really cool. I think it was last year's program at the ARC Industry Forum was focused on smart cities, so I know that there there's a movement out there to focus on this. Could you talk to us about how the smart sending movement, which is more of a macro kind of mentality, is linked back to smart or smarter manufacturing, which is often like the plant-to-plant level. What's the connection there for you?
TL: I got brought into a lot of this Smart Cities initiative as kind of a matchmaker, I would say. Technologists seeing an opportunity to work with cities or county governments or state government, but not necessarily knowing the sales approach and how to call on them. And then on the flip side, some of the people within government agencies not necessarily knowing the technology that's out there. So I've played a bit of match-maker in that aspect.
As I started getting into it, a lot of the cities’ issues are the same as manufacturing, right? How do I do things faster, cheaper, safer, more sustainably going forward? Cities are just like industrial systems. Actually, they might be more complex than industry, cause they're systems of systems, but a lot of the same engineering principles apply. As an example, my undergrad is in electrical engineering, and the principles of circuit design are like traffic engineering. A capacitor is similar to a traffic light, an interstate highway or high speed rail is like a high voltage line, etc. The engineering principles are the same, but the difference is the size of what's moving. It's electrons or people. So that’s why I see there's a lot of parallels, and the biggest difference is maybe the commercial model.
PS: You know you're taking me back to a moment when I was attending PI World and you and I were commenting before we started the recording that we've both been out there and had experience with the old PI World event. The San Francisco Wastewater Treatment Plant was very proud to have deployed the PI System to help monitor the various pumps and valves and production processes. It was clear that they had had some sort of grant money or city money to be able to invest in this, and they had the patience also to get it right and sort of work through mistakes during deployments and optimize the system. So I was impressed that a civic body like the wastewater treatment plant could go through some of these projects and cycles of mistakes, where perhaps a (manufacturing) plant could not afford to or especially a smaller plant. These cities are great experimental grounds, so to speak, to try and pilot these technologies and get them right.
TL: We were talking about Dr. Pat Kennedy who started OSIsoft, and he did a huge amount of work with the city of San Leandro in terms of their grid, in terms of water. I don't know if he funded it, but he collaborated very closely with the local city governments out there to help them deploy PI and use it and improve their efficiencies. I think it's pushing, it's just like a you know organizational change, finding a champion or advocate inside and then helping them change in the process and get the engagement.
PS: I'm curious, what's your timeline for the degree you're looking at two years, three years or sort of as it happens?
TL: Probably three years.
PS: Good luck, it sounds fun, makes me want to go back to school myself. I'm going to change topics a little bit here, and Tom mentioned talking about AI a little bit and you presented at the Fluke Xcelerate event about the different types of AI and industry. It's a topic I'm interested in researching and writing about, and Tom snapped a picture of that chart for me and we've been studying it and I think it made a big hit at the event too. Can you talk a little bit about that? And where AI is in industry, and what you think about all that for the future?
TL: I actually started creating that slide for myself to try to be strategic in where I was taking the strategy for CAT. I had to boil down the big nebulous buzzword of AI, and I boiled it down into the tools, the AI tools and where you could use them, and more importantly, map out the data needs. We were looking at the data that we needed to collect from our products or internally or supply chain, what data do we need? The slides evolved over time and people have found it useful to contextualize the buzzword and as technologies change. You guys saw the latest version that I put at the Fluke Reliability conference. I had a lot of people ask for it so I reposted it to my LinkedIn, so if you want to download the slide, it’s out there.
It's interesting that people with all this buzz around ChatGPT is to realize that it's just natural language processing, and to oversimplify, it's a probability model for words. The AI buzz folks, VC funds, are probably choking with me calling it a probability model and simplifying it that much, but it really is. So I think it's just helping people understand it, it's almost just math.
PS: As far as that chart goes, mission accomplished as far as I heard at the event. There was a lot of buzz in the hallways about how that chart helped cut through the buzziness of that word for people. And you really stratified out the layers in the application said, OK, here's what AI means in this context, here's what it means in this context. That part of the goal of the chart would really struck people, it hit home.
TL: Awesome. I'm glad it was helpful. I'm always about sharing experience.
PS: Well, and you mentioned too that some of those flavors of AI were more prevalent than others; the one on the bottom row for example, was the one that was not being used that much. We've got an audience here who are again, primarily asset managers, reliability specialists. How do you see AI penetrating this industry? Is it still innovative and emerging? Is it starting to mature a little bit for some applications?
TL: I think that's a great question. I had some people ask me the last AI tool that I had on that chart was robotic process automation (RPA), which is basically software robotics writing a program that will do repetitive tasks, and not knowing that that tool existed, we weren't using it inside Caterpillar. I hired an intern, a college intern, for three months during the summer and I gave him a RPA tool and I said, “go talk to people about how to automate what they're doing.”
And by the end of three months, I think he had 10 or 15 projects done and probably another 50 in the pipeline, just because people didn't know it existed. Repetitive tests, you think about it in a business, how often those are. That was eye opening and then it just grew from there, right? So that was an easy way to get started. But again, it's knowing that there's that kind of tool in the toolbox.
PS: I'm curious to know how natural language processing and ChatGPT in particular are going to affect job plan creation, because sometimes that can be a hurdle for people to have time to create a good job plan or iterate on previous versions, but ChatGPT may change some of that for people on the planning and scheduling side.
TL: I think it will. At CAT we used, remember IBM's Watson? So yeah, we worked with Watson on a NLP program to help service technicians sort through service magazines and technical literature. They just read it in there, like, how do I repair this on a D3 version 1987, whatever transmission issue. It didn't work very well at the time. I think maybe that would be something useful. My view is maybe a little bit more conservative about AIs, I think it's a tool for a human to be more efficient; I'm not right now saying it's going to replace people completely.
PS: Writers and journalists are hoping the same thing, too. Whatever level the AI is right now, it really is only going to grow from there and get smarter and be able to do more. So we'll see it growing in the future.
TL: I had this discussion with my husband the other day, it's only as good as the data that you're feeding it. So from a journalist’s perspective, and some others like lawyers, you know, this is going to replace lawyers, right, writing briefs. The thing is, it's still spewing garbage too often, so you either have to have a lawyer looking at it afterwards and making sure that it's legally correct, or you've got to go back and you've got to analyze the data and make sure that the data that's using to fund the model is right as well, and it's vast enough. I think for the danger for journalists, is it spews garbage and people don't recognize that it's garbage, or in any application. You’re still going to have an expert say, is it garbage?
PS: There were two articles that I noticed on this topic too. The first was that a reporter from The Guardian was getting requests for an article that he had written. It turns out that ChatGPT had created the article out of thin air, it didn't really exist, even though he was being cited in references for it. And he himself looked on the website to see what he had written, and he finally had to conclude, no, wait a second, ChatGPT made the reference up.
Same thing for a disinformation researcher I follow, that he was playing with ChatGPT and he found the same thing: it was pretty good at synthesizing ideas, but when it came to referencing the source of those ideas, it was really not there yet that. At best it would make up a couple of sources. At worst it was, it just wouldn't cite sources. So there's some work to be done on that side for sure.
TL: Yeah, there there's pros and there's cons I think right now we're at the we're at the hype phase of it. It's being overpromised for the capabilities, but I think examples are coming up to light, people will be more understanding of how to use it.
PS: I think what we've learned is that a lot of it comes down, like you said, what you feed it and that prompt information, and what you're giving it matters a lot, I think in the end. Brave new world.
I'd like to end on really something fun, if you don't mind. Tom told me you were a cyclist and I'm a runner. I know Tom has run marathons, and it certainly important part of my life, and I just wanted you to talk about that and sort of how it supports your career and your personal achievements.
TL: I was a translate and runner but always had a passion for bicycling. I got my first bike when I was nine years old from a flea market and rebuilt it with my grandpa, and I've been in love with cycling ever since. I love technology too, and I guess from a tech perspective the bicycle is a good reminder that sometimes the right tech is low tech, so it keeps me grounded in that aspect, plus keeps me healthy and outside and exercising.
PS: These are not friendly 10 mile rides. These are fairly significant distances you're going too, right?
TL: Yeah, getting into gravel biking lately and just did a 32 mile gravel ride in Florida. I did Iceland, the Rift the first year that they had that bike ride, that was awesome.
PS: Ah, that's exciting. Do you have any big rides planned like cross-state or across multi-states?
TL: I’m going into the Tulsa Tough in the second week of June, and I'm going to do it with my son. It's like a bicycling Jamboree, basically. It starts Friday, Saturday, and Sunday, and my son and I are going to do Saturday and Sunday, and then the husband and the girlfriend are going to join us on Sunday, so it'll be a weekend about biking.