Why predictive maintenance projects don't scale - with Nick Leeder

Trend Detection Podcast ·

Show notes

Welcome to the Trend Detection podcast, brought to you by Senseye Predictive Maintenance – the platform which enables predictive maintenance at scale across all of your assets, across all of your plants.

In this episode you will learn about:

Scaling Predictive Maintenance: Discover why many predictive maintenance projects fail to scale and how to overcome common barriers.

Expert Insights: Nick Leader shares 30+ years of expertise in manufacturing, helping businesses improve efficiency and competitive advantage.

Big Challenges, Real Solutions: Dive into real-world issues like downtime, legacy systems, and cultural barriers—and learn how to tackle them.

Start Small, Scale Smart: Learn why starting with critical assets is the key to successful predictive maintenance deployment.

Cultural & Financial Alignment: Explore the impact of leadership, collaboration, and real-world ROI on scaling predictive maintenance.

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

Follow Nick Leeder on LinkedIn:

https://www.linkedin.com/in/redeel/

Transcript

So I, I think to start with, you know, everybody loves to have the idea of scale, but it's about who we're dealing with and who we're talking to often has a very big bearing in that conversation. So if you're speaking with the maintenance technicians and teams that are actually doing the job, the boots on the ground, their world is really related to their plant. And the context of scale for them is about scaling the assets they are measuring and monitoring within that plant. So I'm trying to define what scale means. But you, for example, if I'm speaking to a chief operating officer that's responsible for 5 plants or 20 plants, scale to them means actually deploying the same solution across multiple plants and increasing their visibility of the health and performance of those plants. Welcome to the trend detection podcast powered by Siemens, and now your host, Nel Sullivan. Hello and welcome to the latest episode of the trend detection podcast. Um, I'm really pleased to be joined today, um, by another, um, external guest, an external expert, Nick Leader, who's actually, um, an advisor and coach to a lot of manufacturing businesses, and we've. Um, has a lot of insights to share, you know, uh, and particularly around the topic today that we're talking about, which is all about, um, why predictive maintenance projects don't scale. And anyone who listens to this podcast on a regular basis will know, um, we, we've covered this at lots of different angles as well. It's always a big challenge. And I think from a customer's perspective, it's sort of knowing where to start, when's the right time to scale, what they need to think about, um, and all those things we're gonna talk about today. But I guess before we dive into the topic, I'll just like hand over to you, Nick, just to sort of introduce yourself and we'll go, go into it. So over to you. Brilliant, well, look, thank you for the opportunity today to, to speak with you around this topic. Um, you know, my background is really, uh, 30 odd years, which is really worrying actually when I start looking at the amount of gray hairs on my head. Um, working in manufacturing businesses, uh, exclusively on the, the transformation of operations in engineering and in manufacturing, um, and I've done that across the world for big names like SKF and Dyson, uh, and Phillips, uh, but also for mid-sized businesses as well that are seeking to, to, uh, gain competitive advantage or improve operating efficiency, so. I spend my days brilliantly um and I really enjoy it every week, getting out to manufacturing businesses and understanding what they're doing and seeing how all these buzzwords we deal with on a daily basis in the world of digital actually have something meaningful to their organization. Excellent, excellent, and I guess just to, you know, before we go into sort of the, you know, particularly about, you know, the scaling of of predictive maintenance projects, but, you know, talking, you mentioned some big names there as well, but what are some of the sort of the key challenges that they're sort of facing at the moment when you're having these conversations, what are they coming to you to for advice about? So I, I, I mean that's probably. The beauty of my job, it is very, very varied, so, um, you know, ultimately, most people need to get stuff out the door on time, at cost, at quality and do that safely as well, that, that, that last bit often gets missed in that whole conversation. Um, so, you know, some of my topics are generally around, you know, businesses sitting back and saying, look, we've got a really aggressive growth plan over the next 2 to 5 years. How do we do that? Do we need to build a new factory? Can we improve, uh, performance? How does digital play in that? And digital doesn't have to be clever, you know, I, I've got a current, um, uh, newsletter series called The Manufacturing Mouse Trap. You know, there's, there's over 4400 patents for mouse traps in the US. I don't I mean if you don't know that. Only one works really well. I've seen some crazy ones with guns trying to, you know, a mouse eating the cheese and a gun shooting them. But um in reality, it's a finding about how digital gives them simple solutions to their operational problems today. Now that could be a new ERP or an MRP system, and some of my customers have much more broader business challenges which they help them take them on the next journey. It could be an implementation of a mess system on a production shop floor, or it could be, you know, bringing it back to the context of today's conversation. About how do you stop machines breaking on the shop floor, you know, and that has that downtime and that frustration that appears in the maintenance teams because they can't keep the kit running to the you know, the operations director or the operations manager that's not getting the product out the door when they need to because of lack of availability of equipment. So it could be something about, well, how do you implement a a condition-based monitoring or a predictive maintenance solution that helps you get that visibility earlier. So it is, you know, absolutely a broad spectrum of problems that manufacturers face on a daily basis. And sometimes just helping them think, see the simple things away from all the noise we see in the press and the literature on a daily basis of how we can actually get to value quickly. Exactly, and, and today, today we're talking about scale, um, and I I guess to really start that conversation, and we, we chatted just before, you know, we pressed record a little bit about this, but, you know, in terms of ambition and speaking about predictive maintenance in particular, is there. You know, do you see there's an ambition to sort of scale and make this sort of an enterprise sort of deployment, or is it sort of more reluctance in the market, you know, more, you know, they're still a bit unsure about the benefits predicting maintenance and so want to start a bit smaller. What's your sort of feeling based on your sort of experience? Uh, I, I mean, look, we, we all live in the world of budgets for a start, so aspiration and budgets are two different things. So I, I think to start with, you know, everybody loves to have the idea of scale. But it's about who we're dealing with and who we're talking to often has a very big bearing in that conversation. So if you're speaking with the maintenance technicians and teams that are actually doing the job, the boots on the ground. Their world is really related to their plant. And the context of scale for them is about scaling the assets they are measuring and monitoring within that plant. So I'm trying to define what scale means. But you, for example, if I'm speaking to a chief operating officer that's responsible for 5 plants or 20 plants, scale to them means actually deploying the same solution across multiple plants and increasing their visibility of the health and performance of those plants. So scale to different people means different things. But there are, I guess, a number of really important things that are consistent, irrespective of whether you're talking within your own plant in your facility or within um multiple plants and your operations of the company. You know, the first one is, it's got to make financial sense. You know, deploying stuff, I, I, I see, you know, I get involved in questions around condition-based monitoring and predictive maintenance quite a lot for companies. And the first tendency is to try and measure everything. And actually starting with, well look, what assets are actually the most critical assets that when they go down, your line goes down, you can't produce. Or when they go down, you know, it has an impact, but actually it's a very expensive piece of, you know, asset or equipment and actually a long lead time item to get a replacement and you don't want to hold it in stock because that drives your stock levels up and cost and in working Capital. So making sure that there's actually a really clear benefit case of. Which assets do I really need to measure that make the biggest difference? And then start saying, well, look. If uh if this goes down, the cost of production and you know, and we can we all know the business case behind that, but one of the things I spend a lot of time coaching both leaders and people who are trying to get these projects off the ground and to scale. Is, you know, the mathematics of these things you can demonstrate. I recently. Uh, read a paper and actually been engaging with the author, uh, from Standard and Poorer, uh, S&P on, uh, the financial benefits of, um, uh, predictive maintenance, and it was built on a hypothetical example. And their hypothetical example was 50% of the machines in the, the factory need monitoring. But then the example was uh an OEM manufacturer of pumps. Now I'm going, hold on, what, what equipment is critical to you, you know, uh, you know, cos typically a lot of predictive maintenance is based on rotating assets and equipment, yeah, your, your compressors, your motors, your gearboxes. Um, you can do oil, temperature, humidity, all these things come together, but really, do 50% of your machines need to be measured? You know, and you know, the ROI ended up being about 5 microseconds, you know, you turn it on. So let's be honest about it, let's be real about it and actually bring the real case studies to the discussion. So not a theoretical, mathematical case, but. You know what, last week we had compressor number 4 on that line go down and we couldn't produce for uh. You know, 6 hours, so we lost 6 out of an 8 hour shift because of that. That costs you money, that's real. And the reason why it went down was X, Y, and Z. Now if we had condition-based monitoring, we would have known about this 2 weeks, 2 weeks before. And we could have actually planned an intervention in a planned downtime window or over the weekend that would have avoided that. And you know what, that compressor, that's an example last week. In the last year, we've had 14 of those events on those same bits of equipment. Now I'm making these numbers off the top of my head. But what you've done is taken an Excel spreadsheet that gives you some very impressive numbers into something that is real, that is pertinent to the people that want to do it, so that ROI becomes really clear, not just in metrics and, you know, for the finance guys, but actually examples of where it goes wrong. So that to me would be one, a great, a great example of getting it over the line, um. I mean, there are a couple of other areas, you know, we also live in the world of legacy equipment. Um, and, uh, please don't shoot me here, but a lot of salespeople that go in around predictive maintenance kind of come into this kind of view that, you know, you're working on a greenfield site and everything's latest and the, the most shiniest equipment, you just plug things in and off you go. That isn't the world that most manufacturers live in, you know, and 85%, a study by IDC last year said that 85% of manufacturers find that data silos and integration with legacy equipment. Are the major barriers to take technologies like predictive maintenance forward. So, um, you know, understanding that this stuff actually has to be retrofitted, finding companies that can do that with you, both in terms of the technology, but understanding how to um sensorize the equipment in the right places in the right ways have really big impacts. Um, Finally, I think, I think, well, there's two other points actually, I'm probably taking a long time with this scaling question, but these are the things where are, are really things that ignite the scale question is, is also understanding the skills, so. If you speak to a lot of kind of uh we're going into the world of buzzwords for a certain at the moment, you know, predictive maintenance, and a lot of the focus of those conversations I hear are around kind of AI skills, yeah, we need data scientists to do this stuff. Um, I disagree with that because if the tools are doing the job properly. You don't need to be an AI expert, but what you need is the skills to get the data, so the spectra analysis that you get off, um, say, predictive maintenance. And how do you interpret that? How do you take that and say, you know, you're classical CAP 2, CAT 3 vibration analysts. Being able to take that and go, you know what, you've got a flat spot on one of the rolling elements or you've got an issue with the lubrication regime or the shaft's gone out of alignment. So it's about the skills I think that really make a difference are the people who can interpret the data and put them into meaningful um outcomes and that's actually a rarity. So you know, the other reason is you don't have people to really use the systems to the greatest advantage and that we'll see if that has a knock-on effect on terms of the ROI question. Um, And if we're all honest as well, the final one is just culture. Um, we have in the world of maintenance, it has been a world of brake fix or planned maintenance. The vast majority of manufacturers are still kind of fundamentally working in that world. And you know, for some businesses and some sizes and complexity of processes, that actually is OK. But actually if you want to go predictive maintenance around critical assets within your production facility and across your production facilities, which we obviously talk about scale. Then you actually really need to start changing the way of working, start saying, you know what, we're not gonna wait until something breaks. We're gonna proactively get in there and deal with it and the best window and the earliest window we have, uh, and that, you know, from a, from a business point of view, improves obviously your availability of assets, has a knock-on effect on OEE and and you know, and the product you can get out the door. Um, but actually it has another really important knock-on effect of that change of, uh, you know, that get people seeing the benefit of getting in earlier. It also has, you know, maintenance teams feel better because actually fewer things break because you're getting ahead of the game, not reacting to the situation. Um, and all, you know, the stress levels change and you know, it's a quite, it's a really nice, and I've seen this several times when predictive maintenance works really well. The whole philosophy of the maintenance team changes from the guys that are running around like headless chickens trying to keep things up and running to the teams that are actually keeping and building and improving, you know, and today stupid, you know, it's not stupid, but you know, there is energy costs are going up again. You know, well maintained equipment uses less energy. There's a financial benefit in the business and you're increasing the life of those assets because you're maintaining them well rather than waiting till they break. So that, you know, your Capital costs go down and actually for many cases, and this is a real kind of hint in terms of the business case, there's a massive impact in sustainability agenda there as well. Mhm. Yeah, and it, and it's interesting you talk about the broader use cases cos I think. As a, as a market it's moved much more beyond, you know, let's look at reducing unplanned downtime, and that used to be the only metric, and maybe it's still a key metric, but even talking to colleagues on this podcast, even things like health and safety, you know, you know, and obviously you can't put a price on that, clearly, but even that's been talked about, you know, with customers as like something actually we, you know, it could, cos it can help with that if you're proactive about fixing things, it, you know, it's. You can sort of save any sort of risky, you know, repairs that may need to, may need to happen. Well, I mean operating in a safe, in a safe operating environment is obviously number one, you know, quality comes two when it when compared to safety, um, and you know, so we can't argue that, but I think one of the things that predictive maintenance does is it takes the pressure off those moments of failure, when those moments of failure. People are rushing and often when people are put under pressurized situations. It's, it's unfortunate mistakes get made in those, and so it does change that dynamic quite significantly. No, absolutely, and that, that makes a lot of sense. Well, I never really thought about it that way. Um, but some of the, what you were saying before, and I know you apologize for, you know, talking a lot, but A, we're on a podcast, and B, the less I talk and the more you talk, I think it's better for everyone, but there was so much in there, I mean. The, uh, the thing about culture, you know, I wanna just start there because that's something that often, we often talk about, um, Vince Siemens, as well as such a key, you know, key part of scalability, but you also mentioned sort of a technical aspect as well. So, I think Nick, from your point of view, how do you sort of bring those two very opposing sides together and maybe you could, we could even talk about some of the stakeholders involved. Cos there's very contrasting stakeholders involved in, in a predictive maintenance project, and how do, how does that all stay aligned with different priorities? It's a big question, I know. I'd love to. Yeah, I'm trying to, I'm trying to mentally unpack that right now. Um, we could pack, we can break it down a bit, we can break it down. Let's, um, well, let's, let's start at the top, shall we, because it's always easier to start at the top, um. I think a lot of executives view predictive maintenance as a cost in the business. Uh, Rather than an investment and a cost saving opportunity, um. So I, the, the first thing of cultural change at the top is, is the ability to effectively connect how predictive maintenance affects their KPIs in their world. And they're different to the maintenance guys. So having that, you know, it's kind of linked to the ROI question, but ROI's financially measured, but at the end of the day is like we talked about safety a minute ago. So can you say to your operations director, you know what, if we do this, it will improve the safety on the shop floor. That's something he and a clear why, again back to that simple example that illustrates the point rather than the metrics. Those things will stick in their head, and that will start driving that leadership buy-in to do something different. So, let's, we, we talked about the top, and then let's get down to the, the kind of the boots on the ground and actually the people who do the job, so the maintenance engineers and the technicians. How does that change their world as well, you know, and a part of that is, you know, I, I've covered some of that already, is about changing their day. The day doesn't become a constant firefighting, it, it, it's, it's about actually getting ahead of the game, keeping things running smoothly, that pressure levels drop, um. You know, in reality there's a, there's again PWC reporter did a report last year, um, and I'm reading off the screen because quite often I can't memorize all these stats off the top of my head, but, uh, about 25% of, only 25% of maintenance teams use data to make decisions. Wow, yeah, that's quite, that's really quite low. Now that's a reflection on that kind of planned or break fixed type approach, waiting until something breaks. Maybe the data is just nothing's running, but um, when you look at that sort of, those sort of data points, it's that shift from reactive to proactive. That has a profound effect from a cultural buy-in, um, but it's, you know, it's not about the technology, it's about how it changes the day for maintenance engineer, um, and also I think. I, I guess part of that is also about the training you give them. …

Full transcript (WebVTT)