Reliability Gang Podcast
Welcome the #Reliabilitygang Podcast! I would like to welcome you all to my reliability journey. I am passionate about reliability and I want to share as much as I can with everyone with my experiences. Stories are powerful and my aim of this outlet is to gather as many insights and experiences and share them with the world. Thanks for joining the #reliabilitygang.
Reliability Gang Podcast
IS AI TRULY INTELLIGENT FOR MAINTENANCE
Join us for an enlightening conversation with Bertrand from Acoem, who expains how the Falcon Vibration analyser uses Accurex technology. Imagine diagnosing machine defects without needing historical data—this is what Accurex achieves through its innovative use of AI and vibration measurements. Broadcasting straight from France, Bertrand offers an insider's look at how Accurex transforms vibration data collection, especially when historical data trends are unavailable, captivating us with its ability to identify a bearing defects and fault conditions from one reading with advanced diagnostics.
We also explore the fusion of human expertise and AI in the realm of vibration diagnostics. With a robust database of over 50,000 expert-verified diagnostic advice from real use experts in the field this database has been fused into the falcon. We dive into the deployment of these powerful capabilities in Edge devices, ensuring machine diagnostics are both immediate and independent of cloud services. This segment also raises thought-provoking questions about the future of machine diagnostics and the dynamic interplay between advancing AI technologies and essential human insight.
Our discussion takes an exciting turn as we examine the broader implications of Accurex technology. Envision a future where even non-experts can make informed maintenance decisions with confidence, particularly for simpler machinery like pumps and fans. While Accurex is already reshaping industrial maintenance, Bertrand shares insights into the challenges and future prospects of tackling more complex machines. We wrap things up by highlighting how Nest Vision aimed at simplifying access to critical industrial health data, and reflect on the invaluable role of user feedback in our ongoing quest for improvement. This episode is a must-listen for anyone interested in the future of AI-driven reliability management.
Hello and welcome back to another episode of the Reliability Gang podcast.
Speaker 2:We have an extreme special this week. As you can see, we're not in our usual setups and I'm joined by Bertrand from AcoM. How are you keeping Bertrand? Hey, hi, will, hi everyone. It's a pleasure to be here. Me and Bertrand have known each other for a while over the phone. We've had lots of conversations, but this week is the week that we've actually come to visit and we are actually in France right now. We've visited the last two days. It's been amazing how much information and knowledge we've understood about the products and everything, and it's actually been an incredible experience. And it wouldn't be right if we left this whole scenario without us doing a podcast with Bertrand, because Bertrand is the product manager aren't you at QM?
Speaker 1:yeah, actually I'm head of the product team for reliability division. Yeah, at QM, and, of course, pleasure to have you here Will it's a pleasure just in France. Yeah, the opportunity to disclose more about and discuss more about artificial intelligence and how we see that at QM and how this can help our customer.
Speaker 2:What we are fighting for every day to make this happen, and that's the thing we're out in the field using, you know, the Falcons, and we're obviously using it as an expert point of view. And when we first received the Falcons, there was this incredible technology called AccuRex and it was very mystical when we first got it. What is this reading? This is the first of its kind that has an analyzer that kind of tells you what the problem could be, and it was very new and, I'll be honest, at the time when we kind of first were introduced to it, we didn't understand actually how it got to the readings or what it did.
Speaker 2:The thing is, the last two years, ai is a huge thing now on the market. It's kind of now a bit of a buzzword, but when I was speaking to to bertrand the other day, he said well, no, will, this is. This has been around for a very long time and we've been implementing it at aqua for ages, and that's the thing as well, what we need to understand, I think as well, with with this type of technology. It's been around for a long time, hasn't it, bertrand?
Speaker 1:yeah, actually, I mean we released the Falcon in 2013 with a building artificial intelligence for automatic diagnosis diagnostics based on vibration measurements. Yeah, it's actually not the first try for Aquem, it's a fourth generation of automatic diagnostic system and at the time already, I mean it was not so much yet about artificial intelligence or everybody was more talking about big data at the time, maybe machine learning and so on but it was more in the optic where everybody was going to say let us just pump or suck all the data you have and we'll learn over it, do some magic, get back to you after some time. And it's not. I mean, it was not really. This vision was not really suitable for our application.
Speaker 1:We wanted to provide a solution that empowers people in the field with a device we can provide to them instantly, based on a single measurement. Especially when you talk about portable data collection, you don't have years of history of the machine you don't have trends. You are monitoring not one critical machine. You are monitoring hundreds, thousands of assets. You don't have that history on that amount of machines. So we needed to provide a solution that works instantly, based on a single measurement.
Speaker 2:And I think that's the thing. What confused me initially about AccuRx is how could it be so accurate without having again? We've been taught within vibration analysis for years that we're looking for changes over time, we're looking at trends, we're looking at history, and this is for me where it was really interesting to see how actually AccuRex works from a one-spot reading to be able to take into consideration all of that machine characteristics. But where does it pull the data from? And this is kind of where I started to have a great chat with you yesterday and my understanding of how this works is actually it's phenomenal. It's like blowing my mind about exactly how this works because, I'll be honest, I was a little skeptical at first.
Speaker 1:Um, I think a lot of people probably were, and everybody's at the first use, unless you have used it and, of course, yeah, acknowledged by yourself.
Speaker 2:I'll be honest, yeah when I first used it, we did actually have a bearing defect and this was when we first received the Falcons and it's an intriguing test. So we obviously set up a route with Acurex with all the readings and we had our own readings and identified the bearing defect to a tee. And that's when it got my attention because I was like, okay, this is very interesting that that particular setup had no history, didn't have any readings behind it, it was a spot check reading and I was like, okay, this is, this can be very powerful. So we know a lot of ai and data models really usually pull from a source of information, a bit like chat gbt. It kind of looks at you, ask it a question and it pulls all the information from the internet and then tries to give you an answer.
Speaker 2:How does AccuRex work in that sense? Because obviously, where is it pulling that information from? And explain to me. We chatted about Bayesian networks and all of that and this is all very new to me in terms of who I am, but it made a lot of sense. But what I'll let you do is explain how that works for AccuRx and how it kind of comes to them.
Speaker 1:Conclusions Actually, again, the world or the name of the technology is not what is important. What is important is we have a problem we have no history on machine.
Speaker 1:We need to put artificial intelligence in edge on the device, or actually in the device Right in the field, and what do we use as available technologies in the field of artificial intelligence to make that happen? And actually there are two answers. There is how it was at the release of the product and what is the solution. We have designed so that we can improve and be in a continuous improvement loop. So at Aquem, we have been in the field of vibration analysis and reliability for like the past 50 years Wow, a long time, the field of vibration analysis and reliability for like the past 50 years Wow A long time.
Speaker 1:And to release the Falcon with the accurate built-in artificial intelligence, at the beginning we had modellised our expert knowledge through an artificial intelligence solution. So how does it work? It's like a doctor for rotating machinery. When you go to the doctor, you don't know what you have. Okay, the doctor is going to ask you, looking for symptoms Do you have headache? Do you have fever? If I touch here, is it painful? And AccuRex is just going to do the same on the vibration looking for symptoms in the vibration, based on the knowledge we have acquired from our experience. And this was modellised. So more kind of rule-based, so to say, approach based on our expert knowledge, but modellised through a probabilistic approach and sorry for the technical term through a Bayesian network at the beginning.
Speaker 2:But I think it's quite important that people understand what a Bayesian network is, because it might sound quite technical, but what it actually does follows a very different model to traditional kind of you know systems, where it's pulling from a source of information. The information that you guys are pulling from are from years of your own experience and expert information that's being put into your own model. Yes, and then that can only be built over time.
Speaker 1:That's something that can't be just yes, appear you know, we had obviously hours of not brainstorming but discussion between the best experts of of aqua and more, coming from all across the room to agree on what was going to to be the best uh, the best way to make a vibration diagnostic on this type of machine for this type of defect, and so on, trying to to put that into that model that we put in edge, into the system.
Speaker 1:But again, that model that we built was initially based on our expertise, yes, and now we come to how do we improve from that and over the years, of course, we have accumulated more data and we try to say how do you know, actually, how good is your AI? So we had initially a set of data, let's say hundreds of machines, that have been qualified manually by our experts to release that on the market and we had very good feedback and customers would be the best to speak about how good it is today. But then, in order to enter into a loop of continuous improvement, and once you are feeling okay, this machine, maybe it's not the outcome I had expected how do you know that the change you are going to bring to your model is going to be good for everybody else in the system?
Speaker 2:Yeah, I see what you're saying, so you have to make sure that the improvements that are made are going to be able to be sustainable for the future for everyone. Yes, absolutely Moving forward Wow.
Speaker 1:So we have put in place metrics to evaluate and score the performance of our AI, but also put in place solutions so that we can actually collect data from our experts all around the world. Cat3, cat4 certified experts.
Speaker 2:Wow, so you're actually getting that network from people that is then integrated into that network. That allows it to improve itself over time as well.
Speaker 1:Yes, so in essence, we have. I mean, the trick was also how do you extract the expert diagnostic from the data? Historically, I mean, you have been doing vibration analysis also for quite some time and you know the outcome it's actionable data in the form of a report somewhere, so words. And how do you extract automatically information from just text? Just information, essentially Just text. So we tried text mining things, with mitigated success, let's say so. We had to put in place something different, and that's where, in 2020, we have put a new diagnostic matrix which is a way the expert can enter his diagnostic through a codified way which fault I have on which bearing of the machine, which severity, how confident I am in my diagnostic and we have a fully automatic way to codify our expert's diagnostic and then to immediately compare it with the AI scoring. So when we do a change to our model, we know instantly if it's going to be good or bad for the whole population of machines and not just one single machine on which we were trying to improve the model.
Speaker 2:Wow, that's absolutely incredible when you really think about how deep that goes in terms of how you're able to use the expert knowledge.
Speaker 1:But the information, the actionable information, from that input back into probability as well, from that model, that is phenomenal, that is very impressive and going a step further, because first we were still on this, let's say, expertise model approach, where we make some improvements, we are able to score, but now we are at a point where we have more than 50,000 labels, qualified diagnostics by AquamExpert and that is just growing and growing and growing which is growing and growing day after day and we have been able to use different AI technologies to build the model, not based on our expert knowledge that can say right on a paper this is how I make a diagnostic of a bearing fault but just using the AI to build the model for them.
Speaker 1:based on just the database we have. We have this 50 000 plus qualified diagnostic where we know we have a lot of cases with machines with, for example, bearing problems. So we are able to learn from the AI. What is a bearing problem? What does a bearing problem look like from the vibration? Not just based on inputs from our expertise, but looking at way more things than a human eye would be able to look at and I think this is where where we're going with this and where this evolves.
Speaker 2:so, with all that being said, bertrand, now people can understand kind of the model that aqua have been building for all this time and maybe it's one more thing to complete of course, yeah, no, go for anything to interject, so so here I'm talking about learning, you know.
Speaker 1:But it's learn, learning from this database that we're probably unique in the world that we have at AQM, with this plus 50K, 1,000 diagnostic from certified experts. So it's learning from this model. But actually we are putting this model in edge again in our devices, in the tablets in the Falcon.
Speaker 1:So every single technician that has a Falcon and it doesn't need to learn. It's still the same concept that was released in 2013. It just works with a single measurement but with all the power of this model built on our expertise across multiple applications, multiple machines, countries all around the world. This is all put right edge in the device with the Falcon, in post-processing, in our Nest software, with Eagle wireless sensor, with our real-time system, mvx, and this is working instantly based on a single measurement. We don't need to connect to the cloud to learn over a period of time.
Speaker 1:This is all again still working instantly. Thanks to this model, we have put an edge in our systems.
Speaker 2:Wow, that's absolutely incredible. So where does this go from here? Because right now, obviously, as time goes on, as more information gets fed into that, to the Falcon, into the Edge network, this only gets better. Well, what is the future of this now? Where? Where can this go in terms of that real? Should we all worry that we're going to lose our jobs, as we are? Is that correct? It's going to be better than what we are. How do we? How do we now utilize this for the future, which is extremely exciting because I love technology, and where it can go? But what does it mean for the next alliteration of senses and information? Where do you see this evolving into?
Speaker 1:yeah, actually, I mean when we released that in the beginning. So we were talking mainly to our traditional customer users of vibration analysis, all uh needing more time to do what they have to do. So we are looking for solutions that can optimize their productivity, and especially here, just by filtering all machines that you don't have to analyze with AI. That's already a huge step, but now, looking at the performance we have of the solution, we can really consider it as a tool to empower people in the field. Yeah, I see. And when you think I mean you are a reliability expert, how many things can go wrong from the time of the repair to the time the machine is commissioned?
Speaker 2:oh, god, yeah, loads of things. We've seen that happen all the time in in the field of what we're doing and, as well, I mean it's the empowerment as well. Where I see this being really, really effective is that. And the one thing I loved always about aqua and products as well that you've always thought about the expert, of course, people like me, service providers that go in and utilize this very powerful tool, and we can set up our own templates and we've kind of really use it to the way we want to. But one beautiful thing about understanding how this accurate works now as well is that it can be potentially picked up by someone who hasn't got a massive amount of experience and utilise all that knowledge and understanding on plant with a spot check reading. And this is where what we're trying to do at Maintain and working with Aquem as well, is kind of demystify the complexity. Let us do the complex things so you guys can get the results.
Speaker 2:And I think that's always where VA's always been a little bit difficult, because a lot of other collectors we've used in the past have been great for an expert and I have used a range of different equipment, but when it comes down to training someone who hasn't had the massive amount of experience. It's very difficult. They need a lot of time, they need to understand what the readings mean, and this is where I think accorex is going to become extremely powerful, and it doesn't actually necessarily have to be in a falcon either, does it? We can now start to do that on different models as well, can't we so? For me, I think that's where I see this technology going, helping out the people on plant that don't necessarily have the full knowledge exactly, and they can make some good, informed maintenance decisions around what they're doing so not only again with our classic like, let's say, a rehabilitate engineer.
Speaker 1:Customers you know I mean our customers know how hard it is to invest in people, in training and the risk of having somebody leaving for whatever reason. So true, so here with the AccuRex, you have a guarantee that your reliability program is going to provide a minimum performance output. Yeah, but additionally and this is where I was bringing you we can really now think it's not only the reliability people who are going to be able to do a vibration diagnostic, because you don't need to be a vibration expert to know if a machine has a problem or not. Of course, and especially thinking about this, from repair to commissioning, what can go wrong from bad repair, bad transport, bad storage, bad installation, poor alignment, lack of lubrication.
Speaker 2:Here you have a tool that you can give to any mechanics just to guarantee that, once you have done this, the full effort, you have invested money and time to make it right, and that when you put back the machine in operation, it is in precision and that's one thing I maintain, that we pride ourselves on that, the tech-solve-improve model, and we've always really tried to push that, because even now us opening a repair shop and being able to do that part is, you're right, that part from when that repair is done. How do we know when it goes onto site that all the relevant things are done? Is it aligned correctly? Has it been put back into even the transport of that device? Has anything happened between that and the idea is what we love to do is be able to verify all of that with our knowledge.
Speaker 2:But what I think here is giving people the opportunities. If someone had a falcon or could tap into that quick spot check measurement in AccuRex, they could almost verify it as they go as well. So they're using it as a tool as well to ensure that once that asset has been installed, the reliability is within the asset and any issue that potentially could identify we can maybe come off the back of that and make sure these things are rectified before that machine goes into operation. So for commissioning, it could be extremely powerful for people that don't quite understand.
Speaker 1:And additionally for spot checking. I mean, you have done your criticality analysis, you know which machines are going to be part of your program, managed probably by experts doing all the analysis and stuff, but then you can have tools that you can put in the hands of people who are there in the field, in the factory. You can have tools that you can put in the hands of people who are there in the field in the factory I love that who can chuck any suspicious machine, as they are empowered to do so with these tools and potentially pick failures and help optimizing maintenance and everything thanks to that. But maybe one other thing to just not demystify or not oversell the Acurex, because we have been fully transparent from the beginning. It doesn't work on all machines you have in the company and our focus was more to say, there are things we can do well with AI and we wanted this artificial intelligence Acurex to be able to work well on, let's say, 80% of the most common assets.
Speaker 2:Common assets, motors, you can find in the factory.
Speaker 1:So it's like electric motors pumps, fans, compressors.
Speaker 2:Very kind of more simplistic machines.
Speaker 1:And then there are other machines, of course, that are more complex, generating shock, naturally, by design, and so on. Of course, we are not there yet. We still need experts. We have all tools to do that, but the idea of this AI in the former concept was really to help our customers to save time and manage automatically the simple machine, of course, and help you focus your time on where you are needed actually.
Speaker 2:Yeah, and I think that's very important to obviously extend on that point there's no limitations in terms of what it can do. So when you are looking at kind of motors or fans, you you've got certain you know frequencies there that accorex is going to be able to deal with quite easily. But when it comes down to compressors maybe reciprocating compressors, where there's natural impacts and very complex machines that you know even the expert analysis don't like to analyze, accorex is the same. It's then dealing with a lot more complex information that is not necessarily going to be able to. Do you see the future? You know, once you make this you know, say the future is five years, 10 years down the line, do you see this being a tool that is able to tackle them complex machines as well?
Speaker 1:I mean, that's what our customers are asking. So of course, we are looking at it and it, yeah, and and again, the technology evolving, the, the new things about ai coming. Yes, we can be optimistic for the future, but again, what's really important is I mean, you can have the best data scientists in the world to be able to score, to make sure that your ai works. You will need this qualified database or trusted source of data where you know that one, a human expert, who's going to be the teacher of your ai somewhere. Yeah, as input, the right result on the, on the vibration. So just having a concept, an algorithm, an ai technology is not enough.
Speaker 2:You, you must have some playground, you must have the people as well to tie it together, which is so important, and this is one thing I do think a lot of people do assume that the computers are taking over and doing everything. You know the computers are here to help the person that has that information to apply into the right area, and that's, I think, a very important thing to alliterate on. Without the human and expert element of all the information that you guys have been taking for a very long period of time even us in the field, you know, that can help towards that. It's a longer period of time that allows us to be here where we are.
Speaker 2:Without any of that history, it's impossible for us to be able to to have things like AccuRex on the market, because, um, it also does go to the fact that this is not an overnight thing. This is something that has been developed over years and years of people I mean yourself, you've been involved in this for years absolutely, yeah, and that's what think for me when we come in. You explain that and we really understood it. That's what kind of blew my mind, because until you understand the background behind how this is actually done, all people see is, oh, it's a reading that can protect things. It's not. It's been built over a long period of time.
Speaker 1:There's a lot of work behind it for it to be able to do what it can do, and again, the fourth generation of automatic diagnostic AI at ICOEM. So it's not our first try really targeting this application, focusing on the fact that you just have a single measurement and you need to be able to deliver data and output and actionable information from that point.
Speaker 2:So we were just touching on AI, we were touching on the kind of um the accurate and touching on how it can advance. And here's another really good question for you, bertrand before we kind of, uh, wrap this up as well with trend history as well, and with the online models that we've got now, that you've got with Eagle and the MVX systems, where we can take multiple data sets over time and incorporate Acurex, where does this now leave us? So we've talked about how powerful the spot checks can be within this network. Now, if we add trend data and kind of repeat readings, how does this make this even more?
Speaker 1:solid.
Speaker 2:Can we now start to analyze the complex systems as well? With this now, with this?
Speaker 1:that's an excellent, excellent question, actually, and in fact, yes, we have more history, we see more evolution and you know, everything about predictive maintenance is about also looking for how things are evolving over the time.
Speaker 1:So there are way more things we can potentially imagine talking about artificial intelligence and, as a matter of fact, yes, we have some works on how can we address more complex machines, maybe in the first step, not just to print a full automatic diagnostic like we are doing today, but already on more complex applications like, for example, stamping presses, you can get a lot of. It's a very complex process, very well bringing some shocks in the signal. So how do you monitor this type of machines? Also, there is a lot of impact from what is the process, condition, operating condition of that machine, and we are able today to work on different technologies, let's say, to be able to detect abnormal evolution, the fact that you have this amount of data, this history, available for one given machine it opens new possibilities and then we can have, let's say, smart alarming detection that we can later couple to.
Speaker 1:when you get those smart alarms and tell me what is the problem with that. So yeah, definitely there's much more things that will come up in the future in that field of application.
Speaker 2:That's amazing In terms of the products as well. I mean, one thing that you have been very involved with is looking at the products at AcoM and seeing how they can be put to market and where their use cases are and where their values are, and we've had a lot of discussions about certain things and the feedback in terms of us as service providers in the industry and all the rest of it. We're extremely excited now as well because obviously probably I'm not sure if you can see behind Bertrand there's a tiny, small little sensor that sits next to the Eagle, which is something that we're quite excited about as well. The whole product offering.
Speaker 2:Now, in terms of looking at the complex systems where you know you're going to have maybe some NVX systems that are able to be able to deal with some of them, real complex issues, you've got the eagle now that can kind of always transcend data now, rather than taking it once per month, once a day or twice a day with stable conditions to get that volume of information back, where do you see dispara now coming in in terms of an entry level kind of view, because we see a lot of customers that want to be able to monitor things a little bit more frequently, if that makes sense, but being able to understand their criticalities and stuff.
Speaker 2:There's not everything that requires to be an mvx or an eagle. Where does the sparring outfit into this in terms of what it is? And we've used it, we've got a demo kit on right now and the data is extremely clean. Um, for a mem sensor as well, and I think a lot of people may be concerned about that information and we're very impressed with that. And where does it fit into that world for you guys and where do you see it going?
Speaker 1:yeah, obviously, and it's good thing you are bringing the full, the full production product range, especially as you again, you have this background of a reliability engineer and you are able to do criticality analysis.
Speaker 1:You will not potentially address every machine with one sensor and technically, yes, one sensor cannot address all the machines, all the problems.
Speaker 1:So, based on your criticality analysis and what is the risk behind that maybe on the production, maintenance, environment, human safety as well you may define that this machine is very critical and based on the type of failure you are going to detect, you need something very reactive and go for MVX or some machines.
Speaker 1:Maybe you can accept to just have a data collected manually every three months and in between you can have machines that are critical that you would like to get data coming in automatically. And, again based on the type of machinery, you could go either for Eagle, which is wireless sensor with high resolution spectrum, yeah, of course, huge bandwidth. So you have the best anticipation when you have a bearing or component that costs several thousands of euros Maybe you want to have as much anticipation as you want and some machines that can be less critical, but you have high volume of them where the Sparrow will be a perfect fit again to get data commanding frequently, having the ability to make this diagnostic and being able to, like you would do with portable data collection, trick your maintenance actions, supply your spare parts, planning of your maintenance.
Speaker 2:And what I really love about, obviously, this involvement with what you guys have been working towards. I know you've been working extremely hard in the background because all the products that have come out and how they all tie into something that is so special. So we've obviously been looking at the web portal. We're now introducing this to some of our customers as well in terms of seeing that information come in. But the beautiful thing about having a portal is that a lot of the monthly VA data that we have been taking that can now be incorporated into the web portal. So when we're starting to introduce some wireless systems or mvx systems or whatever it may be in terms of criticality, what the customer now can see is all of that information coming into one place.
Speaker 2:I think presentation of information has always been very difficult with service providers and va, because I just feel like dashboards haven't really been kind of. Again, information that we take is important, but the way it's presented sometimes to certain customers is even more vitally important in terms of how is that information digested and how is it readily available. We've had a complete kind of look at the new alliterations to Webport and how it's becoming more accessible. Um, and the beautiful thing is, what you guys are looking at now is trying to input external things back into the web portal as well, so there's a place as well for that. How does that evolve and how does that? How can that place start to actually not just look at vibration analysis information? How can that start to look at reliability and other maintenance things as a whole?
Speaker 1:Yeah, so actually a very good point and thanks for bringing that on the table. But the concept of the web portal is that, as a customer industrial player, you need to have a place where, somewhere every morning, you arrive in the office and you connect to a platform and you can get a holistic view of what is the health status of your factory, generally speaking. And this is what the web portal is going to bring to you An offer.
Speaker 1:An offer. Not only you will get all data coming from the full machines, the full instruments of Aquarium range, the full machines of the full instruments of Aquium range, so maybe portable with a tablet, with a Falcon, wireless with a Sparrow Eagle, or online data coming from MVX system. Everything is into the same software and the web portal provides you access to actually the actionable information you are looking at to evaluate the risk and to plan your maintenance. But on the top, like you said, it's not going to limit to what we can offer. The idea is really to have this portal, this reliability portal, where you arrive in the morning you connect just to one place and you use that one thing, not just different pieces of software, just one place. You can get this holistic view, even if you have some data coming from oil analysis, ultrasound, that in one place you get this and that's beautiful, that that you know that's being evolved and that open source can be able to put in information, because there's nothing worse.
Speaker 2:It's like even in your own life. Say, for example, you've got three or four different bank accounts or investments in different areas. You can't keep track of everything. You have to keep logins for separate things, and you know personally. What we're trying to do is create a place as well where people can go to actually access all that information and then use that to advantage in terms of a customizable way, and a lot of the new additions that we've seen to this web portal allow that to be done. And again, this is just the start of what you're trying to build and grow into.
Speaker 2:Anything else that we can expect in terms of the future? Is there anything that you Again? The thing is, I'm always trying to pull Bertrand's leg and just say what's coming. I can see it in his eyes. He wants to tell me, but there's certain things he's not allowed, which I totally understand as well. But is there anything else in the pipeline that we can expect any hints? Yes, you can expect. Okay, that's all we're gonna get, that's all we need to know.
Speaker 1:That's all we need. But we have been talking mostly about ai today and and uh, yes, as I said, we are going to have a a big release of of the Acurex improvement during the year. Yeah, that's amazing, all our products based on this model that has been learned over our qualified database from 50 plus K 1,000 label diagnosis from Cal3, cal4 experts that we just again put just in the edge in the device that you can use instantly without accessing to the cloud. So that's going to kind of evolve this year as well.
Speaker 2:Yes, we, as well as a you know service provider in the uk, are now starting to really kind of show people what this can be. So we've got great examples, you know, across the uk to show you guys how all of this information can be harnessed into one place. If you'd like a demonstration on that, just just let us know. We're more than happy to be able to show you some of these product ranges and the beautiful thing I I really love about and this is where we're so aligned, bertrand, because we've been talking for the last two days and we've had a beautiful time in france. I'm not gonna lie, I honestly feel like the you guys have. The hospitality has been amazing. We've seen the beautiful city of leon. We've, uh, you know, met so many key people here that you know when, when you're working so hard at home and you've got, you're on the grind and you're doing so many things, you know, when we have these relationships, as you know, it's very business, it's very yeah, we need to get this and what does this mean. But when we've come here, we've really understood kind of all the hard work that you guys put behind all of these products as well. So it's really opened up our eyes and moving forward into, you know, a few into the future, now, from this point in time, that we can really see how like there's a product for for every kind of element of what is required when it comes down to vibration analysis and condition monitoring on the whole, and that's not easy to be able to provide in terms of saying, right, well, let's have a look at the criticality and, regardless of what it is, there's something to offer on every single level and it all goes into a singular place.
Speaker 2:So you know, with this type of new technology and having all these products that are integrated as well, it's not just even a place where we can come and be service providers and do it for you. There's new options, like the Bearing Defender, where it emp powers you guys to go out and take some of this information that can still feed into that same portal that allows us to be able to do the analysis as well and, across the cross, verify a lot of these things in incorporation with acurex as well, if that's something that needs to be looked at. So this is why I get so excited about certain things, because, don't get me wrong, I love products. I love them, but it's the solutions they bring behind that, and I generally think that aqua web have been tirelessly working to that solution, and this is why we're so aligned with the visions and goals moving forward.
Speaker 2:So I just want to really thank you, bertrand, for all the time the last two days we've, uh, had a lot of information and everything like that as well, and, um, thank you so much for your hospitality and everything. Anything else you'd like to say before we kind of wrap up?
Speaker 1:you're very welcome and again, thank you for coming here and having this exchange, and for us, it's very important also to consolidate this feedback of our daily users of our solution. Yeah, we are bringing again, investing in continuous, continuously improving this solution, and your feedback has are very important to that matter well, I don't think this is going to be the last one of these podcasts.
Speaker 2:I do believe that we need to follow up on a series of some of these, whether that's even remotely. I mean, it's good to do it in person. Really, I love this kind of interaction that we do it in person, but there's so much more things along the way, and what we want to be able to is update you guys on, obviously, the products that we use at aquam as well, and, uh, yeah, it's an exciting partnership moving forward as well. So, guys, I just want to thank everyone for tuning in. Uh, thank you for this special thank you and uh, we'll catch up with you guys next week. Take care, thank you, take care.
