Manufacturing is currently moving out of the time of COVID crisis, and many plants report that they are increasing their digital/automation budgets. Kevin Starr works in the digital service space within ABB for North America and, in particular, the process industries component, and has been involved with every step of service and reliability for the past 34 years. Plant Services Chief Editor Thomas Wilk had the chance to talk with Kevin recently about the digital changes currently taking place in industry, and the extent to which they are being driven by COVID considerations.
PS: Can you tell us a bit about yourself and what you've been working on with ABB?
KS: I get to work in the digital service space within ABB for North America and, in particular, the process industries component, but I've worked with ABB for 34 years. I have been involved with every step of service and reliability, and making sure that the equipment that our customers purchase runs and exceeds their expectations. Production, quality, and cost to produce are the world that we live in, and if our services don't help with that, we shouldn't be delivering them. We're finding that that's the direction that the industry is going, and we need to do more of that, and that's what I get the chance to do.
PS: We’re seeing that too on the plant side. The more we survey our readers about how they're using their maintenance and reliability data and who they're sharing it with, we're seeing them increasingly reach out to partners who can help them make sense of the data they're collecting both remote and on-site.
KS: It is kind of fascinating. I was around before distributed control systems, when there were single loop controllers on walls, and at that point, getting information was very difficult. We had a whole series of solutions for manually collecting and connecting, and wires were everywhere, and we were paid to go in and collect data. Then, throughout this journey of DCS and data historians, the data is so everywhere, that we're now getting asked if we can come in and read the data for our customers. It's kind of interesting that the cycle is kind of full circle.
We're using the data, filtering the data tools and the data analytics to do more prescriptive and predictive indications to reduce or eliminate first-time fixes. In other words, why do we have to wait for a failure before we invoke action? You know, we don't have to wait.
That's been the biggest thing for customers and ABB alike (to overcome): "If it's broken, I know it's broke." Well, you also have downtime, you have cost to produce, you've got backup, you have all these costs associated with unscheduled breaks. There is no reason that you shouldn't embrace digital to help mitigate those risks.
So reliability, the face of it changed. I mean, we're in Industrial Revolution 4.0. I used to think that was kind of just marketing. It's not. We're in it. COVID didn't start it, but it certainly sped it up, so it's an exciting time to be in service.
PS: With the rapid adoption of digital technologies, thanks to COVID it feels like right now, in the U.S. at least, we're moving out of the moment of the COVID crisis, and we're moving into a long period of managing the situation. Our readers tell us they're mostly getting back to normal, whatever normal was pre-COVID, when it comes to a balance of predictive and proactive and even reactive maintenance. Can you talk about what you're seeing with challenges getting people onto sites, managing absences due to COVID, and how that's driven this digital adoption?
KS: This idea of conventional service was, basically, you had a guy or a gal with a clipboard running around and doing inspections of equipment, and we call that preventive maintenance or rounds. Then when COVID hit, you can't send people to the site, so now you can't keep up with your preventive maintenance. When there's an emergency, customers say, "I need this fixed, but you can't come on-site."
Luckily, we had started the digital transformation prior to COVID and already had the protocols in place, but it's like anything, people weren't ready for those. We have things called remote insights and some of the wearables and some of the “I see what you see.” We were getting requests almost overnight: "Well, you can't come to site, but I got this thing, this iPad, can we show you (the asset)?" And we said, "Well, sure, you can."
And as I said, (the pandemic) sped up the adoption of digital techniques. We started doing what we called “digital rounds” where we could actually harvest the information. You know, the buzzwords of machine learning, artificial intelligence, they're just philosophies of, “I had a failure, can you look at your data and come up with some model to project or predict that failure in the future?” It's what we always do. That's what we do in the conventional space.
We now have analytics that can build off of data models to provide insight into failures so that we can start scheduling(work): “I know you're going to be down, and the plant doesn't have anybody on it. We can send in a person that is from your region that is safe and is infection-free, and they can come in and do the work.”
We have seen a shift, a big shift towards embracing the concepts of digital, which is to basically automate or project conventional methods into the digital space so that we can react faster. I mean, they don't have to wait for the person on site. Digital never sleeps, and there's no walls in digital, but there's still a P in digital, and that P is people. I can't emphasize that enough. A lot of digital folks have failed when they tried to implement solutions that had no people looking at the information. We’re avoiding that mistake.
Listen to the entire interview
PS: Can you talk about the types of decisions that plants have been making during the past year regarding automation specifically? It's a substantive digital investment, and we keep hearing about plants actually no longer planning three years out. They're actually making the investments now in certain things.
KS: So we're seeing marriages of digital opportunities to enhance physical operations, to mitigate the risk of maybe a recurrence of COVID, or some other something. So how do we embrace this digital road? We're seeing hope, we're seeing the lights are coming back on, we're seeing industries adapting.
There's also something else we're seeing – we're seeing cyber attacks. There's bad guys out there and cyber attacks on our systems. Just yesterday in Atlanta, a pipeline got hit, and that reverberates. Cyber used to be kind of a nice thing. Now you can't ignore that at your site, if you don't know when you've done your backups, if you don't do your patch management. Your IT system is great, but your OT system is what keeps the machine running, and if that gets hit, you could be down, and it could cost millions. I'm not trying to scare people, but that's what we're hitting is, it's not something that happens to somebody else now.
The digital arena, as exciting as it is, it pulls in some people that have different than noble intentions. Fortunately there's a whole arsenal of cyber solutions that can help mitigate that risk. That's a big thing that we're seeing, huge thing that we're seeing right now.
PS: That's interesting at the maintenance and reliability conferences that Plant Services covers, a lot of times there will be sessions on cyberattacks, how to develop a good response plan on the assumption that, if it happens, better to be ready than to not be ready. A surprising number of people in these sessions will raise their hand, and they're able to say in public, "Yes, we were hit. Yes, one server got locked down. Yes, we had to rebuild it." As you say, it's more common than people think.
KS: Nobody wants to see that blue screen, or "Install money here to be able to see your screen." You know, that's why I ask if your screen goes down, how often you run and you need to mitigate. And it used to be, that was enough on service if your server fills up, your hard drive fills up, your controls are all in manual. Now, if someone throws in a virus and it shuts your machine down, that's where we're at, is just the evolution. I'm seeing a recoil, (where plants say), "Oh, well, we can't have any connections from the outside." You know, that's just not going to work. You have to put in the security to make sure you're protected.
The other thing that we're seeing, is that in the traditional approach, we had subject matter experts on a particular asset class, and they were experts at a facility or a region. (Now) the automation space has grown exponentially, so one person can't possibly know the entire stack. What's been happening when there's a failure or a break, the customer says, ''Hey, my system is down.'' When I get that call, I figure, "Well, that could be the power supply, that could be the network traffic, that could be that someone bumped the consistency transmitter..."
What we're seeing is the interconnectivity in the stack of issues that can cause a system to go down are no longer supported by one skillset. And so what's happening is you're seeing workforce balance. You're seeing stress. People are like, "If I get called, I don't know what to fix." And we're seeing the opportunity to help with digital.
You have digital components that can ascertain the health of individual components, but apply analytics to basically expand the sphere of influence of our guys and gals to be able to focus on “there's something wrong here,” and then be able to pull in the experts from all over the world. We actually call that the connected engineer program. That's a big deal.
I think even our customers are like, "If I hire an IT guy, that means I can't get a control person. If I hire a control person, I can't get a chemical person." And so the idea of hiring a point solution, our clients are running into the same thing. (We say), "Well, what if I gave you 10% of all the best?" Really? We can do that with digital. You can't do that with a person, but with people. Nobody can know everything in the digital space. But you can be alerted to areas to focus and that's where I'm excited about, is pulling these things together.
PS: That's a really fascinating insight, especially in the way it's going to change the makeup, I think, of maintenance and reliability teams. Once in a while, I'll talk with people, and we jokingly talk about the death of the maintenance team, for this reason, is that maintenance teams will continue to exist, but not “jack of all trades” maintenance team, because you simply can't afford that anymore. There's no way to handle the whole stack.
KS: Typically what happens with service providers like ABB is, you typically only sell to the area that that person knows, and the plants are massive, so you're leaving areas of the facility completely unprotected because, “oh this guy knows this.” That's what's hard for us as humans, is to realize you can't know everything, but you have to depend (on others).
Where the next wave is going to come, is in mitigation of false positives. And what I mean by that is if your analytics present a failure that's not really a failure, you (might) dispatch somebody and (find out) it's not broken, it was something else. Or (it might be) a true negative, which is everything's fine, but it just broke. We’re already moving into is alarm rationalization, triggers, best practices, golden batch, where should you operate? Where's your sensitivity curve? Those are the next level of things we’re already gearing up for.
That's the other thing is when our limited staff goes and fixes a problem that's not a problem, that's wasted effort. One of our catchphrases is, "The right person at the right time with the right solution.” That's the right value. And that's not just with one person, it's a combination of people, technologies and processes.
PS: We've covered sort of the data collection and user side. We've covered the analytics side with services providers. Final question for today: what role can the system integrator community play in this new industrial normal to help tie up things together?
KS: We're finding it's different by different industries. I've had the opportunity with ABB to work in pretty much every industry that we have, and what we're seeing is some of our oil, gas, and chemical customers use EPCs for integration and system integration, and they're leading the pack.
We're dealing with giant systems that have people working from around the world. It's not just, “you put it in a server, and you have a factory acceptance test.” You have hundreds of people working on a project, and you have system integrators that are integrally involved with that, and they're distributed. We have taken advantage of the digital age to consolidate with things like digital twins, with cloud computing and edge computing, so that people can work from remote locations.
What we're seeing is it reduces startup time, which is primarily what these integrators are after. And what I have to bring up is, a lot of times, system integrators are done when the project is done, but the people on-site, your reliability team, are now owning that. Well, now that's great, they just got something they had no idea (what it is), and which somebody else (developed). In the digital space, we can help with that continuum so that they're not stuck with, "Well, what's that?" Digital adoption, adaptive execution, bringing people together in a collaborative environment – that's what's exciting about digital.
As nasty as COVID has been, is it's forced us to change the way we do business. It's changed our customers. I know it's a scary world out there. We have helped clients plan assessments to help them identify their levels of maturity: Where are they on the digital continuum? We start where they're at with the equipment that they have and move in the (right) direction. Every industry is a little bit different, but we have solution sets that apply across the board. The underlying theme is production quality or cost to produce. If our customers can't make a profitable product, they don't need our help. So our problem is making sure they do that better than anybody. And we happen to sell great automation equipment, too.