Send us Fan Mail Welcome to a special series of podcasts brought straight from Mainstream UK Summit, we are proud to be the official podcast sponsors of such a incredible networking event. AI does not come with a crystal ball, and pretending it does is exactly how maintenance teams get disappointed. We sit down with Nal Solomon from Siemens to get specific about what an AI-powered predictive maintenance platform is meant to do: not replace engineers, not spit out fantasy failure dates, but help real people make better calls with the time and data they actually have. We break down how Senseye (often called SenSi in conversation) pulls in live machine condition monitoring data like vibration and temperature, adds maintenance history from CMMS records, and keeps the human in the loop so the context is right. Instead of treating every pump or motor as identical, the platform builds a unique “fingerprint” for each asset and watches for meaningful deviation. When something shifts, it raises a case, explains what it is seeing, suggests likely causes, and helps teams decide whether to monitor, intervene, or plan work into a future shutdown. Then we go after the messy part of digital transformation in maintenance: confusion in the predictive maintenance market, AI slapped on everything, and the adoption barriers that kill value even when the tech is solid. We talk stakeholder alignment, the IT versus maintenance gap, why reactive cultures struggle to trust insights, and what successful scaling looks like, including an example of a site building internal capability to onboard assets and train others. If you care about asset reliability, condition monitoring, industrial AI, and reducing unplanned downtime without hype, press play. Subscribe, share this with your maintenance team, and leave a review with your biggest predictive maintenance misconception. Support the show
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
SPEAKER_01Hello, ReliaBely Gang. Welcome back to another special episode of the Mainstream Podcast. And this is podcast number three, and we're joined by Nal Solomon from Siemens. How are you keeping, Nal? Very good. Thank you. Very good.
SPEAKER_00Thanks for inviting me on, Will. You're welcome, mate. Enjoying the show so far. Yeah, it's brilliant. Lots of great speakers, and I think um, well, the a thief, but I know there's lots of fantastic people here to network with as well.
SPEAKER_01So it's so great to bring everyone together again. It is nice to be able to bring so many industry professionals under one roof. It's been incredibly good for networking and everything. So, Nal, explain what you do. Obviously, you work for Siemens, you know, explain kind of uh your mission at the minute and what you're doing at Siemens and kind of what you're looking to achieve and you know some of the stuff around what you're kind of doing.
SPEAKER_00Yeah, so I guess first of all, everyone knows Siemens already, so I probably don't need to slightly big name. But my yeah, my role is I lead um marketing um for a product called Sensei, um, which is a predictive maintenance platform, AI powered, of course. AI before it was cool, I always say. Yeah. Um so um, but I've been leading that for around well, uh it's over five years now. But um but Siemens acquired Sensei formerly a like a startup um in June 22, um, and became sort of fully part of it in April 23.
SPEAKER_00Um and yeah, it is um I mean I'm incredibly biased, but it's an absolute fantastic, it's a fantastic it's a fantastic product. I mean, it's provided I can talk a lot about the detail of it, but it provides so much value um to to customers in across all sorts of um industries. There's lots of use cases I go into, but most importantly, it's um it's helping to advance forward predictor maintenance in this world of AI we're in today, I guess, more than anything. Of course. So is it so what is it?
SPEAKER_01Is it a software? Is it a sensor? Is it what is it? Is it a package? How does it work from kind of like user-based to value? Do you know what I mean? What is it? Is it actually a physical thing or is it more of a software thing, or is it both, you know?
SPEAKER_00Well, it I guess an interesting question. So before acquisition, you know, it's we're it was a software company, right? So it's a software product, um, you know, access from the cloud, etc. etc. But obviously with Siemens now, we've got um a much broader proposition where we can include things like sensors and connectivity, you know, a full stack offering. Yeah. Um but the I guess the key to to SenSite is, you know, because I appreciate that not everyone will have Siemens um equipment in their in the in their plants, but you don't actually need Siemens equipment, it integrates with an ETL data sort.
SPEAKER_00And that's really the value where it, you know, which it brings to Siemens, but obviously to to customers as well, because they can um yeah, their investments don't need, you know, they can pick the right products for them, whether it's Siemens or otherwise, and then use SenSi to take advantage or yeah, take advantage of the sort of data and um and the actionable data from that. Okay, I see.
SPEAKER_01So it's uh it's a platform that allows to pull in data and information. So can you pull any information into that, or is it predominantly more vibration information, or what type of information is it bringing in?
SPEAKER_00Yeah, so it's it it I guess maybe to I mean context is key to all this. So it's data from the machine. So then that could be that could be vibration, it could be temperature, it could be talk. Um, it's um for maintenance records. So, you know, obviously we know it, you know, using some of these CMMS platforms, you know, collecting data from from you know years gone by, you can bring that into the platform for context about what the machine's doing. Um, and also it's also the the human in the loop as well, of course. So the the maintenance people themselves who know their machines probably better than anyone.
SPEAKER_00So you combine these things together and you get a uh you know um what we call like a fingerprint on the machine. Because essentially what it's doing, it's it treats every individual machine as uh as a different machine, even if they're exactly the same machine in a plant in Japan, a plant in the US, but they're completely different because obviously what what the climate could be different or one's not level and all this kind of thing, and that's the way it's able to scale as well. So it's not okay. So it's not just certain machine types, it's not just you know, for a couple of machine, your best your most critical machines, but the idea is you scale it across all of your machines and all of your in all of your plants eventually.
SPEAKER_00Okay, I see.
SPEAKER_01So basically it's it's a platform that allows you to be able to pull certain data streams from different assets that can be maintenance history or anything that's happened on that asset for the you know last five years, for example, pull in real condition monitoring data, I take it as well. Yeah, live live data. Live data, so whether that's current like vibration or the rest of it. And once it pulls it in, what is the output? Does it tell you what maintenance needs to be done with that particular asset? Is that kind of the concept, or is it a predictive tool that allows you to be able to identify when something's gonna fail before it does more a CM technique?
SPEAKER_01Or how how I'm trying to distinguish between the difference between a system that just detects problems and something that potentially uses them them AI functions to pull that information together to be a little bit more decisive. Does that make sense?
SPEAKER_00No, no, it does. It does. And so the yeah, the way it works. So I talked about this fingerprint. So if Sensai creates a fingerprint of the machine, so what's the current state of the machine? And it's continuously monitoring that. And then when it sees any sort of deviation or you know, goes a refresher, you know, the temperature rises, you know, over a certain period of time. Maybe it depends on what sort of thresholds are set. But at that point, what what it does, it raises a case, which is just pointing attention to something essentially. So you might you might have thousands of machines or assets, and obviously you can't maintain all of them or go around all yeah, yeah.
SPEAKER_00And the idea is like you should today when so when a maintenance person logs in, they can go, I'm gonna focus on these five cases today. Yeah, and have a look at what potentially could be go wrong before it actually goes wrong, basically. And then and then the case in natural language form provides you know information about you know, we think these are the possible causes. You know, you can see the chart on the chart what's happening here, and this is what you could do about it, and then that's where the you know the maintenance person will go and investigate, um, and they'll either choose to you know leave it to monitor, like we think it's okay, let me let's monitor it, yeah, or they'll uh or it'll need you know more, you know, quicker intervention, or they can plan, they can plan so they'll know okay, so maybe in two months' time we'll need to replace this, and then they can rather than you know, reactively, you know, with downtime and stuff, unplanned downtime and things like that, they can proactively plan it in to whenever they you know they they want to you know stop and have a you know put time into to maintain okay.
SPEAKER_01So how does the AI function work with it? Does it how does it kind of generate kind of what to do? Does it require a lot of data and a lot of history to be able to make like sound recommendations? Because one thing that we do here in our industry quite a lot is buzzwords about AI and how can it work, and we've got this AI tool. And when you really kind of break it down, it's been this is not AI what what I'd saw, this is just a band alarm around a certain frequency range telling you it could be a problem. Um, we know as well that we need to be able to extrapolate a lot of data and a lot of history from our CMMS, you know, and our central maintenance systems to be able to understand like how a machine has operated in the past to be able to understand how it can, you know, potentially fail in the future.
SPEAKER_01How do we do that with limited data sources without having a real context as well? Do you know what I mean? Like, how does how does your software get around that? Because there's always a big big challenge, to be honest, in the industry where if we've got a lot of history, we can use it and we can really benefit from it, but a lot of people are not storing their information correctly now. And if anything, a lot of people haven't even started, do you know what I mean? So, how are we, you know, essentially we're behind the curve of a lot of information and some sites aren't doing it. How do we address that problem, an issue, you know what I mean?
SPEAKER_01With with you know, something like sensei, is any ability to be able to do that a little bit more effectively? Do you know what I mean?
SPEAKER_00Yeah, no, and of course, yeah, no, I definitely recognize recognise that challenge. I think I think the the idea, I mean you have to step back, you know, before we even talk about data when we talk about you know things like that, you sort of have to look at like what's the actual business challenge we're starting, you know, we're we're gonna do that. I mean, because the obvious thing is, you know, for a lot of people, you know, it's unplanned downtime and things like that. But actually what we found with sensei, there's you know, it it's much broader than like I said, you know, with maintenance resources, we know people are leaving um yeah, the industry and it's the struggle to find the skills to require.
SPEAKER_01And actually resource and stuff, yeah, of course.
SPEAKER_00But what I should say on that, it's not a replacement for human beings, right? We're not yeah, this you know, AI again, that that line with AI because this it this isn't like that at all. It's to really help you know make human beings be more productive, it's pointing them to the right. Yeah, more effectiveness, yeah. So, yes, yes, having data and things is is is quite important, but you know, you you have to start somewhere. So it's really so I guess what what I'm saying is it's it you know, don't be sort of oh no, we need like a big bang approach. You could start very small on a very small do a proof of value on you know on the you know, you'd work with the team to decide which machines or assets uh you know for the use case you want to focus on.
SPEAKER_00Yeah, and it's not always the most critical machines because they're often over maintained as well. So you do an exper you know, not an experiment, but you do a proof of value on machines that are over maintained, you're not gonna find anything. Yes, of course, yeah. So you're not gonna find it if you're wrong. And actually, if if Sensai is doing its job, you won't actually find if it was it it can be quite a boring product because it's just you know, you're spotting things, you're you're maintaining things, and then you carry on, and which isn't very, in a way, very exciting.
SPEAKER_01But yeah, I get what you're saying. Sometimes you can almost like you know, eliminate the own you know, eliminate the defects so you don't find a lot, and then sometimes as well. I mean, that's a re-ocurring thing we we find a lot as well, because I I do feel like with reliability and condition monitoring, there is a bit of a conflict of interest essentially, because if you're trying to monitor things, you're trying to find problems, but with reliability, you're actually trying to eliminate them defects so they don't appear in the first place. Does that make sense? Yeah, so you're you're only got this weird conundrum where you're like, right, do you know what? We we need we need to do CM.
SPEAKER_01And I understand that from a plant perspective where they've been very reactive, you're gonna probably find a lot of problems, right? Because you know, they're reacting to the issues and they probably haven't got time to make them right and and you know get get the reliability elements back into the plan. But obviously, at the start of that process, CM's really effective and all the rest of it. But as you start to get less reactive through your time because you're starting to find the problems earlier and they're not failing within kind of production times, you then start to go into the reliability journey of like, right, do you know what? Now we've got our time back, we want to start making the plan a little bit more reliable. We want to start eliminating them defects that initiate, you know, before that then were identified by the CM.
SPEAKER_01But almost when you start to do that reliability model and you start to eliminate the defects, well, then looking on the other side, you don't actually find them as much because the root causes have been eliminated. And then a lot of the time people think, well, what's the point in doing the condition monitoring? So you've got this weird kind of like correlation between reliability and condition monitoring. Do you know what I mean? Yeah, yeah. But how do how do we overcome them kind of contradictions? Because essentially we want to start with the CM process to be able to identify, but we do really want to move into a more reliable world. We don't want to just use our systems to find the problems repeatedly and we're not eliminating them.
SPEAKER_01Does that make sense? Where does the balance lay between the two with this type of stuff?
SPEAKER_00Yeah, and I guess that's yeah, when because when we talk about, you know, and we've we've actually seen like very big successes from like things that Sentai spotted and maintenance that students monitored, and then you know, it it saved them potentially like you know hundreds of hours of downtime, which could cost, you know, lots of money, right? In industries, such things, but that's where you know, where you get to that point where, well, hopefully, I mean that's a wonderful place to be in if you're just sort of optimizing things rather than but but at that point it's about being more efficient, like I said, because you know, you don't know what's going on with those machines.
SPEAKER_00You you can't possibly get around and check manually all of those machines. So it once you've got to the point of you know, stop putting out fires left, right, and centre, you're now beginning to focus on okay, so how can we work more efficiently rather than you know, you know, checking it, you know, checking these machines every, you know, every I don't know, every six months or whatever, you can you can adjust that schedule a bit, and that saves a lot of productivity time and focus, which can go elsewhere. Of course, yeah.
SPEAKER_01You're trying to shift the time a little bit, aren't you, in terms of you know that that reactivity to to proactivity, really, do you know what I mean? Because it's a good trade-off, right? You know what I mean? If we can trade that time for that, then surely in the future, you're you're setting yourself up for success, essentially, aren't you? Do you know what I mean?
SPEAKER_00And yeah, and it's and it's often what we say, because people think AI and stuff, it's like magic, and it'll just one of the common things we hear is oh, so sensor can tell us, you know, it's gonna the machine's gonna fail on this date at this time, which which is clearly not not possible at all.
SPEAKER_01But that's funny you say that now, because it's so true. I think a lot of people look at kind of vibration analysis, predictive technology, IoT, and they're you're right, they think you generally have a crystal ball. But I'm telling you now, if you've got a crystal ball, like can I have that crystal ball? Because like failure is very unpredictable. You know, when we're measuring race weight damage, we can identify there's a defect, we can, and we can measure that as it starts to increase, but the actual failure of that particular bearing can be very different. You know, we did a few studies within our kind of you know, background, we've done a lot of case studies when we're trying to distinguish when does a bearing actually fail and what race weight damage is required for it for the for the cage essentially.
SPEAKER_01Because generally, when a bearing fails, it's the cage that fails that causes the whole failure. And that cage is put under stress from the defect or a spool in the bearing, and as that spool gets bigger and bigger and bigger, eccentricity moves, then there's too much stress on the cage, and the cage then then deteriorates. But we've found even with our studies that when we've found the defect initiate with even an hour race spalling, that the time it takes for the cage to file can be very unpredictable. You know, we've seen bearings you can you don't even need the collector to tell you there's a problem, right?
SPEAKER_01You go in there, it's like everybody knows this bearing is completely naked, right? But when we're looking at that race where we can still see the increase in trend over time, but it kind of gets to the ceiling stage of like, okay, the bearing's really bad now. But the failure rate sometimes continues and continues. And we've seen other you know instances where we've seen a spool, we've we've told the customer to act on it, okay, maybe not as severe as what we found the other one, but even within the next particular surveys failed by that. Do you know what I mean?
SPEAKER_01So it's like this unpredictability at times is really difficult to tell a customer exactly when something's gonna fail. Do you know what I mean? But with with what we've got with bearings, it's just it's it's a real problem within the industry. And a lot of people think you can predict the exact time, and and you gotta look at how a bearing fails, but it doesn't correlate to that kind of analogy. Do you know what I mean?
SPEAKER_00You know, yeah, I think yeah, you you yeah, certain certainly certainly you can't do. I mean, just again thinking, you know, it's when a case is raised in sense, one thing it does do, and again, it's not to suggest suggest it's exactly the same type of failure because every in some ways every failure is different, yeah. But what it can do, what it can do is also compare to all these cases have happened before, and this is the the feedback, this is what happened. So it could, you know, it helps enhance the recommendations. But it's actually happened on a similar type of asset, yeah. We've had the three other cases where it's happened.
SPEAKER_00So and it's it's more of an indication, but again, to sort of close close what I said before, it's the yeah, it yeah, it's not designed to be like this is exactly what happens because ultimately it's up to them the main the human being to do that to do the work, right? Or it's what we describe it as a great description, is like a decision support tool. So predictive maintenance is a cool um a cool industry um term, but it brings up certain predictive, that word predictive brings up certain um uh you know visions in people's heads. But really, I know.
SPEAKER_00Do you like the word predictive? Well, look, funny enough, what I well, five years ago when I started, it was all it was very under, it was still a very underground term because being in marketing, yeah, it was something we've had to educate the market.
SPEAKER_01I don't know if I like the term essentially, because it kind of I think it it leads people to a bit of a false sense of security. Do you know what I mean? Like I get I understand the word predicted in terms of we know there's a problem, but the predictive part kind of almost insinuates this crystal ball analogy, doesn't it? Of like it'll fail on Tuesday, do you know what I mean? Because you know, I just I don't know, it's just it's just a word that for me that is commonly used within our industry, and um, I don't know whether it is really correlating the reality of of of the you know the industry and how failure actually happens, you know what I mean?
SPEAKER_00I think I think it's you know something I've observed over the past five years. Like I said, when I started, predictive was still maintenance was still like this sort of all, you know, people don't really understand it. Yeah, it was a bit of a black heart, wasn't it? Yeah, but but that but now you walk around a you know a big exhibition and it's like predictive maintenance everywhere every yeah I know it's literally buzzword and predictive PDM is like yeah or AI, you know, or AI next to it.
SPEAKER_01But do you think these buzzwords are literally just trying to draw people in and gain attention? I mean, it that is a marketing thing, and we try to do it, you know. Obviously, when we put a podcast out, we'll we'll we'll pick certain words out. Do you know what I mean for the podcast to be able to grab people's attention because of course we want the viewership and all the rest of it, but again, I just don't know whether that's necessarily a healthy thing with this, because sometimes the assumption is that you are the all-seeing eye of everything, and the reality is that is not the case with anything. Do you know what I mean?
SPEAKER_00Yeah, I think so. I think there's a lot of I think the best way to put it is there's a lot of confusion in the market because you know, there's tools that say you do predictive maintenance when they do condition monitoring. There's tools that say predictive maintenance say they're like they're not the same, yeah. Or they're just sort of generic, and that's nothing wrong with this, but more sort of general analytics, right? They're not sp you know what I call something like sensor is like real industrial AI, you know, it's built on you know lots of industrial experience, whereas some are very broad analytics, meant for broad use case, which you can use probably for for some predictive maintenance as well.
SPEAKER_00So and then you've got like you've and when I say that, you've got like small analytics players, and you've got the real big, you know, I don't know, SAP here, aren't they? Yeah, yeah. So you know you've got the big players, yeah, of course. And then and then seems you know, then companies like the more industrial side, you know, like traditional industrial players like like Siemens as well. So yeah, I can see why in the market it's confusing, especially when these terms are used very interchangeably and now AI is slapped on everything. But like I said, Sensai's built on Genitive AI was obviously introduced to the product when you know what was it two years ago, what it whenever it was, but it was is it was built originally whenever it was, you know, early 2010s or whatever, on machine learning, people that you know, which is the like the foundation of all this stuff, yeah, machine learning and those type of techniques, and there's been that's been the platform for many years until obviously genitive AI.
SPEAKER_00Now everyone calls AI everyone just. Yeah, everyone's like the wagon, yeah. Well they put it, they put it in the bucket of AI is just genitive AI, but it's not, it's not, it's it's it's different.
SPEAKER_01It's it's a part of right. What do you do? Obviously, again, I love the technology, I love the premise of it and what it does in terms of you know the the the scene of the future of you know trying to find problems and generate cases and stuff. But what do you do when, for example, you've got this particular software at a business that hasn't quite got the culture to be able to act on the recommendations? Like that, I mean that can be a big problem because I think a lot of people do want this element of kind of technology within their plans, but what do you do to tackle the culture part if it's not marrying? Because again, it's great to have something identify a case, but if no one looks at it and no one does anything with it, like how do you guys deal with that essentially when they're bought into something that isn't necessarily providing the value afterwards?
SPEAKER_00Yeah, I'm really pleased you asked that question. A because we are we are doing a joint report with mainstream, so I'll just plug that um straight away. Yeah, plug it away. Which we're going late later this year on um change management and and the challenges of adoption. And the reason why we sort of commission this um, I say commissioned or working with Mainstream on this report is because we know it's almost a bigger challenge. People talk about um you know technology being the challenge, you know, which technology do we use? And we talked about the confusion in the market, but actually it it's the adoption challenge because again, there's probably you know a thought, you know, in some quarters that you know we buy we buy software, we buy some licenses and hand it over, and then off you go.
SPEAKER_00If only I'll hear my um hear my colleagues um saying to me, but um predictive maintenance is you know, is is you know, when we think about, you know, I think there's been a lot of talks there have been some sessions about you know reactive cultures and stuff, and that's very difficult to go from a culture where you're just waiting for something to fail and then you're it's true, where you've got all this coming.
SPEAKER_01Exactly, because you've not been selled for, have you really?
SPEAKER_00You know, essentially no, but you're going from that to maybe a data you know condition monitoring where you're you know you know starting on that journey to something where you're actually SA trust, which is an important word, you're trusting a platform like SenSai that what it's telling you and what it's where it's guiding you is is the truth.
SPEAKER_01And so So do you need to be a specific place for SenSai to work then? Does that make sense? Because we talk about them culture barriers and it's it's a big problem for us, actually, as a reliability business as well. You know, we we get the same issues, right? We come to a plan, they want to make a change, but they're so early with their, you know, there's such a reactivity, even the stuff that we tell them right now, they've got the time to do because they're fighting the fires that they've got, like, and that's difficult for us because we we know that they need to start somewhere, but it's a bit of a painful journey out of that like kind of remit. Like, is it the same for sensei?
SPEAKER_01Is it difficult to start in that place? Do they need to be at a certain adoption stage with with the ability to be able to act on things, or is it a case of you can start anywhere with it, but it has to be roadmapped, you know?
SPEAKER_00Yeah, it's a it's an interesting one. I think I mean all I can say is so one thing we've we've learned, I say we yeah, we've learned over. The past sort of 10 years, a lot of learning some different deployments and working with different customers in different course, because you've probably got customers that all have different stages as well.
SPEAKER_01I mean, they're not all at the same stage, are they? You know what I mean? Everyone's at very different kind of journeys as well, aren't they? On their journeys, you know.
SPEAKER_00No, exactly. I mean, we we we have, you know, we we have uh we actually have a methodology, right, for a deployment. It's like a five-stage where you know, where everything that's gone into that is and it's been improved, iterated over the years through learnings from previous deployments. So we've learnt that certain we know that certain projects don't work, or if you don't have certain stakeholders involved from the very beginning, that it's not it it's not gonna work. For example, one random example is you know, if someone really wants to do predictive maintenance, but they go, Oh yeah, we'll give it to our intern to have a look at.
SPEAKER_00And we know those projects will fail.
SPEAKER_01Yeah, but they don't have the ability to be able to, you know, comprehend that and actually deliver, do you know what I mean? In terms of the training as well.
SPEAKER_00Well, because you need to coordinate with so different different, you know, reliability engineers and maintenance people and you know, IT and all this guy, you know. We had examples where we've been speaking to the maintenance side or the yeah, and you know, about Saint I was great, great, great, and it goes along, and then uh IT are brought in at the end, and they're like, Well, no, we've got you know our own systems and whatever.
SPEAKER_01So is that what how how is this, you know, the I I I know exactly what you mean because we've been there three or four times as well. With you know, you do the reliability stuff, you do the failure mode stuff, you're like, right, you know, quite complex machinery, here's the way that we monitor in it is quite intricate, we have to do it in a special way. Then, as soon as you actually get them concepts and you're right, IT get involved, for example, they're like, no, we've got our own systems like well, hang on. Like, I understand you have a system, but we've gone through all of the strategies in terms of how we monitor this, and this is the way that it has to be done, and none of that's taken into consideration as well, sometimes.
SPEAKER_01And there's a big gap there. Do you not find between sometimes what IT wants or what maintenance needs? Do you know what I mean? Because we've we've seen that quite a lot, actually. It's quite I'm quite glad you actually raised it. Do you know what I'm saying? Well, of course, because I think it's a big problem that we find you.
SPEAKER_00Well, I mean, that the all yeah, this IT OT sort of um yeah, and no, and uh there quite often is because often you know you could say but Sensi sits on the on the engineering side, and then like I said, you've got these you know bigger, let's say big IT players who we don't know who they are, you know, where you know, and naturally they're like, Well, this is approved and we couldn't we can use this time.
SPEAKER_01But then you've got the users on the ground, let's say the the end users, with another important, you know, mate massively, because I mean that information that they log uh uh on the you know on the ground floor is is so important that it is is accessible and it's usable and it's easy and it's feasible for them to actually put this information in as well. Do you know what I mean? Because without that, you're losing clarity, aren't you, on the main as functions as well, you know?
SPEAKER_00Yeah, I mean you couldn't yeah, because uh because that's the other way, is when and we've all been there. If you work for any any sort of company sometimes where you just say, Well, here's your tool, that's it, right? And that and that that's all you and that and that's all you can do. But um, I know we sort of yeah, we sort of went round, but yeah, yeah, it like I said, if um yeah, it's all very well, you know, since I'm generating these insights and things like that, but if they're not actually that that's coming back to the the original point of your yeah, of your of your question. I mean, if they if they're genuinely not ready, you can't you know for we're not a sort of a change management, you know, organisation or you know, or you know, in in that respect.
SPEAKER_00But what what we can do is support, we know we've worked with hundreds of customers by now, you know, who've been on similar journeys, different points of the journey, but um but where we've got this methodology where we can go, right, we can make sure we select we know the right types of assets to use, we know to focus on the use cases you want to do, and then it's scaling step by step. And we and we have like customer success teams, for example, who will you know have weekly calls to go, okay, so what have we seen in Censor this week?
SPEAKER_01Have you acted on it, you know, in this sort of gentle try almost trying to yeah, you're kind of trying to you know motivate them to be able to do them things as well, aren't you? Because essentially you want them things to happen because that's when the value's driven through the software, right? You know what I mean? When things actually get actioned.
SPEAKER_00And then you then you eventually get to a say, I mean, one customer example I can I can give, um, and it's all on lot on the Siemens website for Blue Scope Steel, where you know they've just scaled and scaled and scaled. I can't remember how many sites, or an Australian um steel company, but they they've got to a stage where it's almost there's still involvement from Siemens, but actually they've taken a lot of stuff on themselves. So they onboard assets themselves, they have a sensi school where they um you know educate other sites around the globe on you know how best to use sense, and that's exactly the position you know that if every customer was like that, that'd be fantastic.
SPEAKER_00Appreciate people who they're probably a special case actually in some respect, in a good in a very good way. In a good way, yeah, of course, yeah, yeah. But um, but yeah, but yeah, so and that just you know, yeah, it it's um yeah, in the ideal world that everyone everyone would be like that. But people start in different ways, but the the point being there's a lot of fantastic guidance from the you know the Siemens team to help, you know, you're not just sort of given a software and off you go, you're sort of guided along the way, you know, how to scale, how to justify the scale internally, how to capture the return on investment on the deployment and this kind of stuff.
SPEAKER_00It's all it's all part of it.
SPEAKER_01Yeah, yeah. That sounds great. No, what a great conversation! Thank you so much for sharing the knowledge, mate. Honestly, it's been really insightful. Um, Sensi seems like a really cool platform, it seems like it's very value-driven as well. And um, yeah, we've got a bit of a running theme today. We've been asked a question from the last podcast. Um, and basically, I want you to also ask the next person the question as well. Because it could be around reliability, it can be around anything, really. Do you know what I mean? It's you know, you might want to throw a bit of curveball in there, Nile, if you want to. But her question was what do you think is the best characteristic of an engineer?
SPEAKER_01Well, that's a good question, actually.
SPEAKER_00It is a good question. Um, one that one that like wants to buy um send side no, no, no, no, no, no, no, no. But on I guess to continue the theme we were talking about. I think it's um it's having sort of an open mindness to new technology. I know it's I know it's I know it's it is a bit crazy, I know, with AI and everything and I mean fantastic sort of products everywhere here today, but I think it it's being open-minded and what the technology can deliver, that it's not going to be a magic bullet, but actually it's there to support your productivity, not to replace you, but to actually enhance it.
SPEAKER_00To support you in hard to, yeah, yeah. And just think of it that way rather than just being closed book. Yeah, like no, no, no, I don't need it.
SPEAKER_01You know, be re-open to it.
SPEAKER_00So what we what is do what we're trying to say is is combining this amazing, you know, 30, 40, whatever years of experience, you know the machines better than everyone, with technologies help you to scale that expertise. Yeah, and it's a powerful combination. So that that's what I would say, and it's all links into what we've been saying. Okay, awesome, Nile. Uh question for the next view for the next podcast E. Well, actually, I want to go to I don't know if anything about AI has been asked so far, but what what I'd say is if nice, if tomorrow someone pressed a button and said AI is not allowed, you know, wiped out of everything, you know, how would you how would you react?
SPEAKER_00How would you react in your you know whether it's personally to you or whether it's in your you know in your role in business? But yeah, if it was completely banned tomorrow, what would you do? Yeah, Nile.
SPEAKER_01Great question. Thank you so much for jumping on, Nal. I wish you uh a good rest of the show. I've been too plenty of people to to go and see, but no, thank you so much for jumping off.
SPEAKER_00It's been a pleasure. Thanks, Will. Take care, Nal. Bye bye.
Send us Fan Mail Welcome to a special series of podcasts brought straight from Mainstream UK Summit, we are proud to be the official podcast sponsors of such a incredible networking event. Reliability gets talked about...
Send us Fan Mail Welcome to a special series of podcasts brought straight from Mainstream UK Summit, we are proud to be the official podcast sponsors of such a incredible networking event. More data does not...
Send us Fan Mail A 25-year engineering skills gap doesn’t get fixed with slogans, and it definitely doesn’t get fixed by buying shiny new tech while the basics fall apart. We’re joined by Wayne Stannett, founder of...
Send us Fan Mail If you’ve ever thought, “We’ve got condition monitoring, we’ve got a CMMS, so why are we still firefighting?”, you’re not alone and you’re not imagining it. We talk about the uncomfortable gap between...
Send us Fan Mail Breakdown culture is expensive, exhausting, and strangely normal in a lot of factories. We wanted a straight-talking conversation about what actually shifts the dial, so we sat down with Dan Bridges to...