How predictive maintenance can enable digital transformation - with Tom Jacques
Show notes
Welcome to the Trend Detection podcast, brought to you by Senseye Predictive Maintenance – which gives you visibility and insights into all your assets, from single machines to full plants to help you reduce downtime, increase knowledge sharing and accelerate digital transformation across your organization.
Why Predictive Maintenance is the Gateway to Digital Transformation Discover how starting with predictive maintenance (PdM) helps organizations shift from reactive to proactive operations—building trust across departments and unlocking long-term digital value.
How PdM Breaks Down Silos Across Teams and Technologies Learn how implementing PdM brings IT and OT together, fosters collaboration between operations, maintenance, and finance, and uncovers valuable insights from both new and existing data.
Real-World Success Stories from Automotive and F&B Industries Hear examples of PdM in action—from saving over £200k at a stamping press to digitizing workflows and creating digital twins in global food and beverage plants.
Scaling PdM Without Overhauling Everything Get practical advice on starting small, using existing infrastructure and sensor data, and building scalable, future-proof PdM strategies that don’t require massive upfront investment.
Debunking Common Misconceptions About PdM Understand why digital transformation doesn’t need to be fully complete before starting PdM—and how this approach actually accelerates integration, ROI, and long-term growth.
You can find out more about how Senseye Predictive Maintenance can reduce unplanned downtime and contribute towards improved sustainability within your manufacturing plants, by visiting: www.siemens.com/senseye-predictive-maintenance
Transcript
I'll tell you, if you're overwhelmed by where to begin with digital transformation, start where failure costs the most, and that's your machines. and predictive maintenance is the lowest hanging fruit with the highest return of investment. Welcome to the trend detection podcast powered by Siemens, and now your host Nel Sullivan. Hello everyone and welcome to the latest episode of the Trends section podcast. Um, today I'm joined once again by my, um, colleague, esteemed colleague Tom. Welcome back to the podcast, thanks for coming back. Hi, thanks for having me. Um, um, I know we'd usually do, um, introductions and things like that, but I think we could dive straight into the topic today because people know you by now, Tom. I mean, you're our. I'm happy to do that you're engineer for Sentai within Siemens, right, unless you wanna add to that in any way. Mainly, mainly I'm solutions engineering, so anything where you think there might be a requirement, um, reach out to me, reach out to the rest of the team. Um, I'm always here. If there's any questions or anything, always feel free. That's awesome. Hopefully I didn't undersell you there either, cos, you know, your role is very, very broad and a lot involved, and that's why we invite you on the podcast, obviously share those, um, thoughts and, and especially the conversations you have with customers and potential customers. But I mean, so today's topic, um, we're talking about predictive maintenance and how it can enable digital transformation. Um, and as we know, digital transformation is such a broad topic. Um, and to us, predicting maintenance is a core focus, but in, in a lot of ways, it's a tiny dot on the map of a bigger project, or maybe it isn't. I mean, that's what we're talking about today, right? So, Let's, I mean, let's get, let's get started and straight into it. I guess if we start with the big picture first and sort of expand a little bit. So, Tom, how would you define digital transformation in the context of industrial operations, first of all, just to lay the, the groundwork for this. Uh, that's a good question. So I'll start by saying it really comes down to changing, so digital transformation means shifting from find and fix to predict and prevent effectively. So it's about connecting assets, unlocking data, making sure that you can do something with that, with a main objective to do things like reducing downtime, you know, improve efficiency, make sure that the factories is a little bit more resilient to things, so you're not just reacting to problems as they arise, you can foresee some of that stuff coming before it comes. And crucially as well, making sure you don't do too much, so a big element of this is definitely, you know, how can we prevent lots of over maintenance. Um, that's the fundamental aspect of predictive maintenance. Digital transformation goes then much more broad and it becomes, you know, what can we have a digital twin version of, um, how can we start to drive this across other areas of the business? And it, it really starts to become now making sure that these different softwares and different solutions integrate with each other. That, that's a big headache for companies at the moment where, where these different tools don't link, um, as well as they could. And do you think sometimes, because I guess when anyone hears the term digital transformation, it can almost be a bit overwhelming because like I said, there's so many different tools, strategies, places to get started. I mean, do you, do you find that from customers that sometimes it's a bit like overwhelming to start with? Um, I think it can be, I think it can seem overwhelming, but really when you start with predictive maintenance, you start by not just, you know, you're not just fixing the machines, you're building trust across different departments. So it can seem daunting, but it becomes then, you know, how do the finance team and the operations team and the maintenance team have this common goal? What's the, what's the common impact that you're trying to, trying to look at? And that, that impact comes from everything. I talk about predictive maintenance because it's my specialty, but it's the same for all different areas of a, of a process. And, yeah, I guess if we, if we focus on the mains, well, that's what, you know, focusing on the maintenance side, and I guess, I mean, different organizations, we're in different, um, stages of their journey, but if we think about, um, the challenges organizations face, even just digitalizing their maintenance strategy, maybe we could talk about that first of all. What are the, what are those challenges? One of the main ones is trying to unlock the data that's there, so many, many machines sit within most of these facilities or mines, quarries, steel processing plants, whatever it could be, and what you end up with are lots of machines that work effectively siloed and maybe there's tags and little pieces of information that say, yep, my part of the process is, is complete, move on to the next step. But there's lots more information within that. There's things like harmonic frequencies, there's pressures, there's flows, and all of this stuff changes over time. Also, the process changes over time. Maybe you start to manufacture a different product on the same pieces of equipment and, and some of that machine will be brand new and some of it will be legacy. And again, you have to figure out the best methods of making it all really work together, um, harmoniously. And that's, that's the difficult part of all of this and also having of course different PLCs from different vendors, making sure all of these things can communicate with each other so that you as a business can do what you do best, making chocolate, making steel, whatever it might be. And that for me is really one of the biggest challenges, making sure that all of these different solutions work together, um, and that is daunting. I think that is a, is a daunting thing. Yeah, I, I guess that's particularly, um, the case, you know, when you talk about, um, I mean, a word we use often, you know, about scale and across the enterprise, as soon as you get, you know, to a deploy digital transformation, I guess, across multiple sites, that's where the complexities start. Like you said, with different systems, different approaches, different culture, maybe as well. That can be part of it, yeah, so if you've always been in a, in a reactive mode, changing, I, again, I will link straight back to predictive maintenance, it becomes a cultural shift to change your approach. And I guess, um, yeah, so let's, I guess, illustrate some of, some of these points, um, we've discussed so far in terms of sort of real life, you know, or real world, um, examples where predictive maintenance has enabled this broader digital transformation. Is there anything you can share, Tom, on that side? There is, the, the, the main one I would go to to start with would be a global automotive plant that Sens and I worked with, um, looked at digitalizing the workflow but kept facing, um, you know, they kept facing costly unplanned downtime, um, on a single stamping press. And just by implementing predictive maintenance on that one asset, they saved over £200,000 and effectively that was good return on investment came very quickly. So that's a good example of where we've done this. And a global food and beverage customer, just thinking of a second, um. With large plants around the world, they weren't really sure where to begin. um, when we talk about digital transformation uh with an example where predictive maintenance it really benefits you, I'll tell you, if you're overwhelmed by where to begin with digital transformation, start where failure costs the most, and that's your machines. And predictive maintenance is the lowest hanging fruit with the highest return of investment. And since I worked to centralize existing sensor data and mapped that data across um into a digital twin of the factory to begin with, locations of where the motors are, the pumps, the conveyors, ventilation systems and so on. Um, and effectively what we did then was we gave them visibility as soon as we could and gave them a digital fingerprint of all of those assets and understood very quickly what, what was healthy behavior and what was unhealthy behavior based on those metrics that we started to create. So, you know, 2 to 5 days normal operation since I build you a fingerprint of normal, of normal healthy operation. And equally it could be in a, in a failed state already, so. That's a question we get asked often, so I'm caveating myself while I say it. Um, a machine could be on boarded in a failed state, and again, every case closure gives you the chance to fill that back in and say, yeah, the machine was already failed, uh, you know, the bearings were 80% collapsed. Pretty much that last little grinder and that increased torque, we replaced it, that's why you've seen different results. But effectively this has helped them to streamline what they do. Uh, maintenance is now much more optimized and, and really it focuses on how much maintenance they do. So that's something that we've really helped them with, uh, and that's just, you know, two of good, two of many examples that I could give to you. Yeah, and I say 22 examples that, um, are very different, different verticals, right, with different approaches. I mean, is there, when we talk about sort of digital transformation or predict maintenance as an enabler for that, you know, what sort of the differences, I mean, compared, you know, automotive versus food and beverage, if you want to put the two up against each other. Is there, Differences, differences in approaches or, or mindset or there is, there is, so if you've got a, a process that runs start to finish nonstop, so if, if you think of something like paper manufacturing, that's something that you can't really stop it, it needs to continuously run, which means when you get the chance to, to stop the machines, typically you'll you'll really strip things down, you'll change bearings that aren't perhaps, you know, completely collapsed, but you wanna make sure they make it through the next window of operation. And that could be months long. Equally, if you've got a process where it stops and starts, that's where you might not need to do so much over maintenance, but reacting to stoppages, all these small little downtime events, conveyors that stop, bearings that um collapse and all the rest of it, they will start to impact you, you know, collectively over time. So in food and beverage, that's a good example. Um, in steel, that's a good example, you, you wanna make sure that downtime is avoided, and I've, I'm picking a few key ones that everybody knows, but there are, of course, hundreds, thousands of different processes, um, and things being manufactured, you know, tens of thousands of different things. What I would say is when we start to look at this, it comes down to what's the operation and, and what's the value of the product that's being manufactured or processed. And that is really what, what starts to impact, um, you know, how much can we spend on the project as an initial outlay, what's the return of investment. That will change depending on, um, you know, the business impact that we have. So the, if something's more valuable, if you start making medicine or you're, you're making, um, vaccines or vials of things or even tablets, this stuff is very valuable. So an hour of downtime here is enormous compared to. You know, making uh potato chips or something smaller uh and, and less, so we have to factor that in and we, we do factor that in when we do projects, um, to make sure it aligns to the, to the industry. No, yeah, it's really interesting, and the thing is, it's an important point to, um, to look at these nuances, because I guess when you're, when you're talking to these different customers and different. Verticals, it becomes clear which use cases are maybe more appropriate to one over the other, right? Yeah, exactly, and it's also understanding, you know, you might get a very forward thinking company, um, the sector is then irrelevant, it's a company that are trying to move forward, uh, and maybe take the risk of, of deploying something that they're not sure of, um, and also we get customers who are a little bit risk averse and. They need a little bit more uh of an understanding first before they effectively dive in at the deep end, you know, this is tried and tested, we can, we can show that, but it's, it sometimes it really requires that um handholding and we've got an approach for, for everybody. And I, I guess another, well, I mean, key to maybe predictive maintenance as well, but certainly to digital transformation, a great deal of collaboration, you know, involved across many different, um, departments. But I guess if we focus on the predictive maintenance aspect, how does sort of, how does implementing predictive maintenance sort of bring, bring the data, the systems and the people together so they can sort of collaborate together and achieve those outcomes, some, some of which you've mentioned already. Um, I would say when you start with predictive maintenance, you're not just fixing the machines, you're, you're building that trust across the different departments. So if it's robot reliability teams, operations, finance, they will all see an impact. Um, and that alignment I think is quite rare across a business to also have teams of people that start a project and they really sit right down at the bottom of the business, um, you know, at the crux of things, making sure the machines do what they should do, all the way through to the CEO. And predictive maintenance, sometimes that's why it's complicated because all these teams suddenly realize they've never had to work this closely together before. Um, but equally then you start to see the results as a team, you start to achieve something, you start to have a direct impact, um, you know, return on investment impact. It makes people's lives simpler if you've, you know, you're not, not running around chasing after yourself so much, you can preempt things, you can. Order spares and, and, and parts into stock so they can be changed at the right time. That will have an impact on the process, the product, so you can make more, you can, you can do more with less. And of course, right the way through the business, that has, uh, knock-on effects, positive knock-on effects all the way to the top. And actually, I wanted to pick out something you said sort of earlier in your, in your answer. You said about, you know, departments, you know, coming together that potentially haven't worked together before. Is that something you, you see quite often, that, um, oh well, A, that, you know, that's the case, and B, that predictive maintenance helps to bring these teams together. Yes, that is exactly what we see. We see companies where it's a new thing, digital transformation, um, in its nature is fairly new. Um, in some organizations, it's already set up to, to work well, uh, maybe whatever it is they manufacture, there's lots of links across the business. But what we also see. Are the ones where they, where they don't, and, and predictive maintenance gets them all in the same room. It makes you start to look at things and say, right, do we have the right equipment? Do we have the right process to deal with something when it goes wrong? If it goes wrong, what do we do? In some cases, like we, like we're saying now, if it's reactive, you, you can use predictive maintenance to flip that whole process. So actually predictive maintenance is a fantastic place to start in a digital transformation journey. Uh, because of how many different impacts it has, and it pays for itself immediately. Um, immediately, could be, depending on the process, a few months, um, a, a few years. But it It depends as well on how much stuff you've already got in place, so if you don't have an existing, you know, network system of, of DCS and SCADA where data can be collected into one central location, saved in a historian. And then sent somewhere, you need to start actually implementing that first. That's one of the things you need to put in place before you go any further, so that that data can be made available, made useful, um, and I guess what it helps you to do is understand what, what gaps have you got in the organization, um, pretty quickly. I think that's what predictive maintenance is good for, that's one of the reasons. I like starting projects with predictive maintenance, um, with an industrial edge solution in place, with the Brownfield Connect solution in place, and some other products from Siemens. Um, fit harmoniously into this. So, I, I guess when there, there is, let's say nothing in place or very little in place, I guess it's about building out a sort of a roadmap, you know, uh, and being, I guess, realistic on timelines and what can be achieved and, and when, but maybe you could talk a little bit about that process, Tom, for sort of mapping out, you know, if we give an example of, um, an organization with very little, um, data collection, etc. in place already. So the essential part is really um to have an element of standardization to start with, just across the process, so your DCS and SCADA network. And then as far as I'm concerned, that's the main standard that we need to pull data uniformly into one place. And often I think that's the part of our projects that we, we really look to deploy a simple Siemens solution. I just mentioned it, so something like Edge, something like Insight Hub, other products from Siemens that are really, really useful to do this. Um, and also from there, it starts to give you the ability to monitor things like energy consumption, look at areas of wastage, look at OEE, um, calculations, using effectively the same information. So digital transformation is, is doing the most you can with what you've already got. Um, and if you've, if you've already got some sensors, what data can we pull from? Do you already have historians? This is part of the scope project that we start every project with. So we understand, first of all, what are the limitations, what are the. Uh, constraints that we already work with and as much as possible, build a solution that fits in, um, without huge Capital expenditure of, um, you know, sensors and additional edge boxes and everything else. As much as possible, we'll use what's already there. But equally, if you're looking to transform, we can put a transformation layer in place, um, to benefit you for the future. That's also got cybersecurity involved. Uh, and all the rest of it. And this is then positioning you very well for long-term digital growth, digital transformation, adding new things, adding different, um, you know, different PLCs from different vendors, different sensors. You know, we, we use different sensors where we need to, um, maybe for a better IP rating or for a, a higher heat tolerance or something like that. And because Sensi is designed as a, uh, an agnostic solution for any industry, any space, and also Brownfield Connect is specifically designed for things like this. So for linking, uh, different types of data together. We end up with a solution which is based on a few foundational technologies but gives you an essential, um, you know. Process to build on for the future and that's, that's, you know, we're, we're involved, um, in many different projects and, and the Siemens Accelerator Project is designed kind of with this in mind. So the, the accelerator portfolio in its entirety is how can we remove paper-based processes and do as much as we can with the data that's there and then simulate things and digital twin things. Uh, this is the direction we will move in, uh, lots of companies now look towards this. …