Seeing the Invisible: From Strobe Lights to Modern Predictive Maintenance - with Richard Ella

Trend Detection Podcast ·

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.

In this episode, we’re joined by Richard Ella, who takes a step back in time to show how some of the most powerful ideas in predictive maintenance aren’t new at all and why understanding their origins is key to explaining, adopting, and trusting AI today.

What you’ll learn in this episode:

Why modern AI‑driven predictive maintenance follows the same principles as earlier mechanical and electrical innovations

How the strobe light was originally invented for maintenance and what it teaches us about “seeing” machines differently

A simple, practical way to explain AI and Senseye without buzzwords or hype

How AI mirrors the instincts of experienced plant operators by detecting subtle changes before failure

Why curiosity, trust, and change management matter more than the technology itself

How early warnings become real value only when teams act on them

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

They know so well when something has changed. They don't know what's changed. But the first indicator that you have for predictive maintenance is that something is different. You want to get there before, you know, things start breaking and, and this, this is what AI is really replicating in the physical world. Welcome to the trend detection podcast powered by Siemens. And now your host, Noel Sullivan. Hello and welcome to the latest episode of the trend section podcast. I'm joined by one of our, our regulars, uh, Rich Nella. I mean, it's been a little while, Rich, but I'm really pleased to, um, have you back on, so welcome back, first of all. No, thank you very much, Noel. It's been a bit of a break, but I'm pretty excited about today's episode. Yeah, and, uh, and you, you're right to sort of say that as well, because we're, we're, we've covered a lot of topics on this podcast, and we've covered them from different angles, but this one feels like a completely new angle, even to what we've done before. And I'm gonna sort of let Richard. Come in and explain it a little bit, but we're sort of looking back at some other technology from, from the past and comparing it to maybe today, like to predictive maintenance and bring it round to that. But I'll let Richard explain it as he sort of brought the topic to the table. So, yeah, well, why did you bring it to the table, Richard? What, what did you want to discuss today? No, thanks again. I, I was actually listening to one of your previous podcasts with a Nick Leeder on what actually works in manufacturing. And it actually got the little cogs in my brain working because one of the things that I do every day, whenever I see new software, if I see a new feature or any new update that we get, I, I always, when I'm walking around a, a, a factory or if I'm just looking at a feature in, in the software, I always ask myself, how did we do this in the past? And curiosity is really what drives me and gets me down these rabbit holes. But what, what is really interesting, and it's very important when we're trying to articulate what Sensii does is draw it back to the origins, pull it back to what did we do prior to AI and, and then attach what we do with AI. And so, if we look at, you know, all of our digital tools and it's a, it's a digital representation of what we used to do physically, mechanically, or electrically. And, you know, as simple as I was caravanning on the weekend and in this little kind of room, uh, out in the sticks, I switched the light on and it was on a five-minute delay and switched off. And I was like, wow, that, that's, that's pretty, pretty advanced to have a 5-minute delay. They feature on a little hut in the middle of the sticks. And then when I actually went and had a look at the switch, all it was was a little vacuum chamber that you pressed. And depending on how quickly the air bled out of it, it would close the switch again. So you can see there that it is such a simple thing. And if you achieve it electrically, or if you achieve it mechanically, you, you know, you, you can do it in many, many ways. And so, What I really wanted to actually talk about today and maybe share a story with you is What I found with the origins of the strobe lights, you know, I know this is a little bit left of center, but stick with me here. You know, I think we've all been in the nightclub and enjoyed the strobe light, 2 o'clock in the morning to see random movements, but. You know, what I actually found when I was looking at the origins of the strobe light. It was actually invented and used for maintenance activities in heavy metal and, and rotating machines. So you can go and have a look at it. But there's a guy, Harold Edgerton, who actually was working on this. And once he had developed a way of taking a picture and flashing at the same time, he was able to capture bullets in motion. And it was a real breakthrough in the photography space. But What Harold did is he actually took, packed his suitcase up, went around to all the factories in the US and he used the application to flash it against rotating fans and rotating equipment so that they can actually do a visual inspection as it's rotating because obviously, if you flash at the same rate as what the equipment is rotating, it looks like it's standing still. And, and I thought, that is absolutely amazing. I had no idea that that the origins of a strobe light were actually used in maintenance. And the, the parallel that I'm trying to pull here is that what the, what Harold was able to do there, he was able to Provide actual a visual insight and look at something in a different way. In his way, he was flashing a light at a, at a rate that matched the equipment and he could see, it looked like it was standing still. So, this was Maybe one of the first ways of um doing maintenance in a different way. You don't need to stop that piece of equipment anymore, because you can now inspect it, the blades and everything while it's rotating. So a real shift there. And The, the parallel that I'm trying to, to explain here is we are doing the same thing today with AI. We are, we are now using different tools. Might be a flashlight. It might be language learning models and machine learning algorithms. But we are doing the same thing in a different way to derive better insights into our equipment. And so, if we look at, and if you break down what Sensii does, for me, you know, it's very important to have that. That elevator pitch on what Sensii does. Because as soon as you start talking about AI everybody glazes over and they say, look, they, they, you know, they want a simple explanation of what a Sensii does. And if I could explain it very simply, AI executes the same function that a very experienced operator of a plant does when he walks around his plant. If you, if you, if you've met a couple of these guys, they, they are so in tune with their environment, the sound, the vibration, the feeling, the smell, the ambient temperature. They know so well when something has changed. They don't know what's changed. But the first indicator that you have for predictive maintenance is that something is different. You want to get there before, you know, things start breaking. And, and this, this is what AI is really replicating in the physical world. It's finding those nuances that you don't normally pick up. And it can do it at scale. And so, be it Harold Edgerton with his flashing light, or be it Sensai looking at your data with the, with the algorithms and everything that are set up to look for those nuances. It's, it's a much easier story than to sell to a customer because they can then understand what it's actually doing because nobody wants to onboard something with thinking it's going to take their job. They, or with all of these generic throwaway lines in the AI world. So, This is what I wanted to talk about today and, and just share an experience of what we used to do in the past is exactly the same today. Except with a flashing light, we are just using machine learning and large language learning models to see what we couldn't see before. And so, I mean, thanks for, first of all, sort of open, opening up the conversation there. I mean, with the, the first thing that sort of, sort of sprung to mind is with the, the stroke, like you mentioned AI and as well as a lot of the world sort of embracing it and, you know, it's wonderful, it's doing amazing things. There's obviously skeptics out there. So, I wonder if he went through a similar thing with the strobe light. Where he's bringing, you say, going round to factories, bringing it in, and there'd be people with a similar mindset, sort of, like, oh, I'm not sure. I guess it's with any new technology, right? There's always gonna be skeptics, but there's always gonna be, you know, the, the early adopters, and we know that sort of curve. But do you, do you see what I mean? I'm sure he went through the same process, right, but in a completely different era. OK, that's right. Could you just imagine somebody knock, uh, knocking on a factory door and saying, listen, you know, I, I can. Uh, look at your equipment while it's rotating, standing still with a flashing light. That, that apprehension is still there today with and that, and the only way that you overcome that is, is to be curious from a customer's point of view, be curious and let us show you. As soon as that customer sees the setup and the flashing light and his fans standing still, he then gets it. That's the challenge that we have today with AI is getting from that concept to an actual, uh, detection or forecasted failure where the customer goes, wow, OK. I, I, I, I, you know, I knew that was going to happen. I didn't think the, the algorithms or the systems that you have in place would have detected that. So, I think it's important to understand that AI is a very much a disruptor. It it's, it's very much changing the way that we do things, but it won't change what we do and how we use it. That will remain the same. We constantly want to drive more value. We, we want to use the latest tools to, to give us better insights. And so, so yeah. Could you just imagine this guy rocking up with, you know, this big case and saying, listen, I can do this for you. And I can imagine he would have been, you know, shown the door a couple of times. But as soon as one customer, uh, demonstrated the, that. You know, it, it would become standard practice. And actually, I actually went on further to ask the question, how much money did Harold Edgerton make out of this? And it was virtually nothing. But it's one of those situations where he was at the forefront of doing something new. It was how you applied it that got adopted in so many different ways that wasn't a direct reflection of these efforts. And, and I think that's what AI is doing today. It is so powerful. And you need to point it with purpose. Otherwise, you just end up with a, you know, a noisy chatbot, you know, that that's, that's pointless. That's just generic AI not focused on task. Whereas you look at what Sensi have done. They've really blended in your machine learning, finding the problem, referencing it to a similar challenge, and give you your conversational like interaction. So, you know, this is definitely the future as far as how we interact with our machinery. And, and, you know, SI are definitely at the forefront of that, where they bring that all together. And you can see many different companies all trying to do, use AI in different ways. And, you know, I, I think AI just like the, the strobe light, it's, it's how you use it and how it gets adopted and translated to value. That kind of determines its, you know, real value. Yeah, interesting, I guess another, another point of it, because we, we mentioned about, you know, this person going into factories, he's probably had a lot of doors shut, but, you know, it only takes one to sort of really, you know, to start, you know, the momentum behind it. But then once, even once you've got someone on board, say, oh, yeah, I'll give this a go, it's a, and I relate this to Sensai as well, but it's, you know, it's, you know, it's actually believing the insight, you know, using the insight and then taking action. With the insight as well. So it's such believing what you're seeing. And in the world of AI that's certainly applicable, but it could certainly be applied to the strobe light and other technologies in the past too, right? Yeah, and look, when it comes to adoption of technology, it's always been around. AI has just super scaled the The resistance barrier. OK. So one of the things that I find very interesting, because I do a lot of work around Australia on mine sites and very much remote locations. And, you know, I'm, I'm very fortunate that I meet and greet a lot of people in the maintenance space. And one of the things that I see very, very quickly is I can see the people that have not I don't want to say the word given up, but they have stopped being curious. All right. And, and usually when I have a conversation with those people, it's very much a particular point in their career, where the technology that they knew and understood changed, and they didn't change with it. And when you don't change with the technology, you get this resistance barrier that comes and it's natural. It's, it's a human nature thing. And so, when we look at technology, you know, we've had so many hardware and software breakthroughs, but we lose people along the way. That's the downside of this kind of AI generation world that we are going into. The people are, are, are, are feeling overwhelmed, and the, it's killing the curiosity. Um, you know, AI. So it's very important to, when, when you do talk about AI. is that you take people on the journey and, and that it empowers them to do their job better. That's what Senseii is all about. It's, it's all about just tapping you on the shoulder saying, listen, I might have found this. What do you think? When you, when it gets adopted like that, and the, the human in the loop feels that they are empowered, that's the perfect scenario. But That transition from doing all of the analytics yourself and doing all of that human analytics and handing that over to something else and trusting it. Takes time. And I would say why you will see my angle here is very different is I will always try and, uh, link what we do today to what we did yesterday. And the, and we are just doing it in a different way. Somebody actually, I was at an event recently and, and one of my colleagues actually said this to a person next to me. And it really hit me is When we talk about innovation, Innovation is not the application of the latest technology, you know, that's at the cutting edge. Innovation is the adoption of existing or past technologies to address tomorrow's problem. And, and that's what you really got to do when you, when you are looking at AI, make sure that whoever you work with has demonstrated the value, has demonstrated what they can do for you, uh, yesterday, and, and, and for the last 5 years, applied to your plants. And, and that's what I think Sensi are very good at, is demonstrating. How they've achieved success in their vertical, whatever it is, and not talking about what we can do and, and up in the air, you know, generic statements. And, and, and that's what opens the door for Harold when he goes to, on his next tour is, he says, listen, uh, are you interested? They might say, no. He said, listen, this customer has saved, you know, 20% downtime. Is that something, not something that Interests you, and yes, that's what you, that's what you really want. Yeah, and that point you made, Richard, about, you know, at a certain point, you, you stop being curious. I mean, I may, yeah, no I guess that's a thing, not just manufacturing, but just any. Walk of life, right? I mean, it, it's a, it's a very general point, but do you think what you were saying there about the proof points, things like that, does that change the curiosity in any way, or is it quite fixed? I mean, let's talk about the manufacturing side, obviously, but is that sort of fixed, or is there any way to sort of encourage and, and get people on side, or, or is that quite an ingrained belief, let's say. I, in my role here in Australia is to assist customers with digital transformation. It's actually got nothing to do about the technology. I can talk about the features and bells and whistles all day long. But if we don't have a company culture that embraces curiosity, Or it has processes in place to stimulate that, this free thinking, this, this ability to, or this bandwidth, I think maybe that's a better word, um, is companies need to understand that you cannot force change. Change happens on its own, and change will only happen when somebody overcomes a problem themselves. And, and so you need to be patient. And, and stimulate the curiosity. Because if, if, if you just stuck to pure process, you do it this way, you don't do it that way, um, that runs dry very, very quickly. And so, but all this being said, I do feel that, that AI has caused a wave of apprehension. It's everybody's got it. And so how do we overcome that is we, we articulate the story. Look, this is just another tool to achieve an outcome. And that's the way to see it. And, and when you adopt that tool for you to do things better. So, you know, it's, I would say it's our biggest challenge. And it's a, it's a, it's a human nature thing. This is nothing to do with You know, technology or anything like that, people won't adopt or change until they actually see the value for themselves. As soon as they see the value, as soon as they go, OK, that shortened my email time by 20 minutes. As soon as they see, I don't have to go out and inspect 50 machines anymore, I'm getting the same accuracy until that is actually achieved. People will always sit on the fence and you can't kick them off the fence because that's, that's not natural. So, very, very interesting topic and maybe we can have another podcast on this one. But I, I'm very interested in And those people that have stopped being curious. I'll challenge them in a, in a, in many different ways. Sometimes I'll be like, aren't you curious about this, that and the other thing, and see if you can bring them back and to, to, to staying curious. So, yeah, that's a, it's, it's a pretty big topic. Yeah, I mean, I guess the phrase we're, we're talking about here is, is change management, right? I'll mention this because funnily enough, I'm actually recording a podcast tomorrow about that exact topic with Nat Ford, or actually I should say, Natt's recording a solo episode, which we've never done on this podcast ever. So, it's gonna be a, a new one. But it, it, but it shows the importance of the topic. I mean, it's, it's something that's always. Raised and the point you made about, you know, not selling technology necessarily, but actually the change required underneath, and I guess our friend with the strobe light went through a, again, went through a similar process, right, because it's a, presenting something like that, not only unknown, but also, You know, potentially a big change in how, how they work and trusting the technology. So again, the parallels are, are clear to see. Yeah, and I think one of the things, there's so many slogans and, and throwaway lines out there, but what I do in my daily role is not focused on the features of our product. I'm focused on, uh, solving their problem, OK? Because that's all a customer is interested in. Yes, you can delve into the how for the geeks and the people that really want to look under the hood. But at, at a, at a high level, you should be able to quickly communicate how you address their problem. So they say to you, all right, uh, you know, we, we failed 20 pumps in the last 10 years on this application. That's what we saw. And SensIi is just a tool to achieve it. And, and, and that's the important part because if you get stuck on trying to sell the technology, the, the, the barrier will come up straight away because you're trying to get them to understand how we solve their problem. And so, it is very, very important that you, you go to that level when, when needed, but you should be quickly be able to communicate, oh, look, I know, we understand that problem. This is how we've solved it in the past with this technology. And they go, oh, OK. I, I get that. And then they'll ask that question and be curious and then go and understand it, um, in time. So, I guess, as we, we sort of drift towards, towards the end, I'm gonna give you sort of an early warning, Richard, something I didn't mention before, and since we last recorded, I've actually introduced a lightning round as well. So 3 sort of quickfire questions. So, we'll do that, but before we do that, I'll do one more, more question. …

Full transcript (WebVTT)