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
VIBRATION ANALYSIS STRATEGY - METHODS BASED ON FAILURE MODE
How can you transform your plant maintenance strategy with the latest in vibration analysis and condition monitoring? Join us as we uncover the secrets behind leveraging wireless sensors to revolutionize maintenance practices. We’ll explore criticality assessments and failure mode analysis to ensure you choose the right technologies for your needs. Plus, meet Will, our new Technical Director, who’s on a mission to balance cutting-edge innovations with cost-effective solutions for clients new to the world of condition monitoring.
We'll also dissect the potential pitfalls of a one-size-fits-all approach to wireless sensor installation. Learn from real-world examples where traditional vibration analysis falls short and discover the importance of meticulous evaluation and planning. We highlight the key differences between the Sparrow and Eagle sensors, discussing their unique advantages and applications in industrial settings. From hazardous environments to the biogas industry, find out which sensor is right for your specific needs.
Finally, we delve into setting the correct frequency for data collection and the move towards proactive maintenance strategies. Understand the benefits of continuous monitoring systems like MVX and how they support plant digitalization. We emphasize the need for cultural and operational shifts to keep pace with technological advancements and preview exciting upcoming projects. Tune in and get inspired by pioneering maintenance strategies and the future of digitalization in industrial plants.
been a lot of thought process for me and I'm thinking what are we going to do a podcast on? Cause it's been a while since we've been doing this with we had main tech and it's just been, I'll be honest, like the response from main tech was incredible. It was just incredible. I literally I've still I'm still obviously in meetings from that from that period and we're still having discovery calls and all sorts of great stuff, which is amazing. But as well, that means that the focus has shifted a little bit. But this week I've just been really kind of thinking about how do we pick this back up.
Speaker 1:It's been a while since we've done a podcast.
Speaker 1:Where are we at with everything and with a lot of the calls and the sales calls that I'm having.
Speaker 1:It's very interesting because obviously when you do have somebody sell calls from main tech, they are talking to other providers as well, because obviously you're not the only one at main tech.
Speaker 1:There's other people that do what we do and all the rest of it, and they're there and you know they're having discussions and they're usually trying to you know, obviously have a discovery who is the best provider provider for them with their aims and goals and everything. So a lot of these meetings that I've had, I've noticed in terms of strategy, especially vibration analysis strategy, in terms of what they're trying to suggest, is that I'm seeing a lot of kind of kind of wireless sensor options being thrown across popular now, a lot more popular in what we're doing and again with um this year, our strategy for that has changed massively as well. With the sparrow we've got a lot of demo kits going out, even some of these customers as well. But as well, what we're trying to also do is kind of you know, obviously it's great to sell some of these sensors and to monitor things, but we as a business need to also understand the reason why you'd want to go for that option, if that makes sense.
Speaker 2:Yeah, so I said nice, I did a linkedin post recently all about the why, why do we put a sparrow on, why would we choose an mvx? And I think that's what we're going to cover today.
Speaker 1:Yeah, definitely it's gonna be an interesting discussion because, as well, in perfect timing, will's role has changed. So, um, this is probably going to be the, the reveal, shall we say. But obviously will, as you know, has been dealing with the operations of the business for the last two, three years, I believe, and now we're moving into a different age. We understand how important that the technical element of wireless sensors, aquem Vision, the data reporting processes through kind of like refined AI as well in terms of how do we utilize that for the business? Will's role has now turned into a technical director role. So this is a perfect time to talk about this, because me and him have had a lot of discussions, because generally, within maintain, my kind of role is looking at the future, like, where's the business going to go? What do we need to do now to ensure, five years from now, that we're in the right position where the industry is going to be? That's always really difficult when we're moving fast, but culture's not moving with it. Does that make sense? Because we're like, oh, we need to be prepared, we need to have all these processes available if we need to do a bit of 4.0, but the problem is when we're getting that stuff ready.
Speaker 1:We're talking to our customers and in and around us and some of them not even doing condition monitoring to begin with. So this is where I feel this conversation is going to add so much value, in the sense of if you're a customer that haven't got any kind of condition monitoring or vibration analysis, how would you set up that strategy? How will we define where you would put some of these wider sensors? Because some, some customers I'm talking to, they've not even had any handheld, they've not had any version of it, and they're now obviously coming into this world and having discussions about wireless sensors. But I think what me and will have to do now is give you guys like almost what are the stages and why would we put online sensors on?
Speaker 1:Where is the best strategy for vibration analysis and the most cost effective one as well? Yeah, because let's be honest, putting a sense on every single asset within a plant that has 500 assets is not going to be cheap and is it necessarily going to give you the value that having a guy coming around and doing some good observations and taking data is going to be. So today's podcast is going to be vibration analysis strategy. So I'm going to let Will talk a little bit about kind of failure mode. Again taps into reliability.
Speaker 2:Again a lot of this does and I think sometimes if you haven't had the reliability studies done on the plant to understand criticality, to understand how something can fail, this task can be a lot more difficult, kind of well yeah, we have to kind of like take a few steps back a bit, because the whole idea of doing condition monitoring is because at some point down the line, somebody generally has said that we need to look at doing a condition-based maintenance approach.
Speaker 2:So a lot of maintenance implants who don't do condition monitoring are going to be reacting to their maintenance, so they're in a reactive maintenance state or they're going to be changing it every so many hours just because they want to. And I think now there's a general consensus that everyone understands that doing condition-based maintenance is a good thing to do, saves money and only does the maintenance when we need to. So initially, the whole idea of doing condition-based maintenance and vibration analysis is that vibration analysis is a very effective condition monitoring technology that can detect a wide range of failure modes very effectively and generally the main ones that we're concerned about bearing defects.
Speaker 1:Yeah, the ones that are going to stop you in your tracks Generally. This is, I think, why vibration analysis probably is the most popular. Is it because it can identify the things that are going to you know, cause your plant to stop?
Speaker 2:Yeah, and in reality and generally speaking, what we'd like to do is we'd like to go to the plant and we'd like to do the criticality assessment and really understand the criticality of the machines, and we'd like to maybe do a FAMICA or do RCM to really understand what maintenance needs to be done based on the ways that potential machine could fail. But not everyone has that tool set available to them or the experience or necessarily initially, the time, because all of that takes quite a lot of time, which is why we often find that vibration analysis, especially in plants that don't do anything, can be a great just let's get going and doing something to get some buy-in as well about the bigger picture of reliability.
Speaker 1:And I think this is why most plants, when they're having failures and it's causing quite a lot of chaos within the plant. This is why a lot of plants do look to do some condition-based maintenance, vibration analysis, off the off, before they've done them studies. It just makes it a little bit more difficult to understand what strategy to apply. But this is where maintain can help, in the sense that we don't do a full-blown criticality assessment, a full-blown Femica and all of these assets. But from our experience of obviously doing this for years and years and years, a lot of the machinery that we're monitoring in terms of these plants are very similar to the places that we have done these studies with, you know.
Speaker 2:And that's the important bit to take from that as well is that we are maintaining reliability. We're reliability engineers, so that our experience and expertise comes in at, not just looking at how we're going to monitor that plant with vibration analysis.
Speaker 1:We're not just vibration data collectors. We're doing the mini famika in our head. We already know, don't we before? Again? Don't get me wrong. There's some real complex machinery that is going to require that, if that makes sense.
Speaker 2:When you're looking at the hidden failures that you're not really sure about.
Speaker 1:I mean the majority of most plants.
Speaker 1:When you've got, you know, the common rotate machinery motors, fans, gearboxes that you know generally run a process of a plan, we already kind of know what is important or not.
Speaker 1:And I think once we've kind of had a look at that criticality and we've kind of done our own kind of assessment and we've had a good conversation with the customer, quite quickly we can get an idea of the assets that probably need to be in this program. The beautiful thing about vibration analysis and handheld vibration analysis and we know this now more than ever from moving away from, you know, older collectors to the aqua and falcon is that we can, on the high speed piece of equipment, take one reading that will have three axes of vibration and it will cover multiple different failure modes in terms of frequency range in 9 seconds for one bearing. So if you can imagine an asset, a fan for example, you're probably going to have 4 points on that fan, 2 on the mower, you know 1 for each fan bearing. You know if you're taking 9 seconds per read and you are literally probably taking data on that asset under 1 minute if it is running and it is obviously in the same process conditions.
Speaker 2:And you have to think as well, with that handheld data collection, which? Or portable data collection, within that one minute that you've data collected, let's say, let's be extensive and say a minute and a half, yeah.
Speaker 1:Within that minute and a half, let's add some more time on that.
Speaker 2:Let's just add some more time. So within a minute and a half an engineer has gone to that machine. They've probably felt the machine, assessed how hot it is. They've probably felt and assessed how much it's vibrating just with their hand. They've tested it with the data collector, they've screened the data, they've understood if it's okay, if it's okay, they've left it. If it's not, they've made notes on it and they've already done the pre-analysis before anything's been done. And that done, the pre-analysis before anything's been done. And that's all been done in a minute and a half. And that is why portable data collection, the visual aspect of it, you know there are failure modes, belts that can wear, that don't necessarily enough to cause the vibration, and this is the other thing as well.
Speaker 1:If a plant hasn't and quite likely if we come in to do condition monitoring and they haven't done the initial part of criticality yeah, failure mode analysis then how do they know that their pms are valuable?
Speaker 1:Probably not valuable in that sense as well. So that also means it's like again, you know, the wireless sensor model you know assumes that that can detect the defects that vibration analysis can pick up. Right, if we haven't done proper failure mode analysis on that particular asset and understood, well, what about the belts? What about all the things that VA can't pick up on the other failure modes, you're going to miss them. And this is the thing with you know. Ideally, if you've done that properly and you've got good PMs and people on plant that are available to do them things, then maybe you can put a wireless system on that and know that it's going to pick up all the failure modes because you've got the other ones covered. The problem is that we're seeing in most of this industry that people are coming in Blanket approach, blanket approach, putting sensors on everything. They've not even asked the customer what other plan may have gotten them.
Speaker 2:We went down to that customer down in devon who are now we've been working with quite a long time and they're now we're putting the aws sensors on. Yeah, and they put it on a bearing. That was what 10 rpm maybe, not even that I think it was a cooker bearing.
Speaker 1:It was on like a. It was yeah, and cook a very we can cook steam cooker right. And they're putting these sense, they're putting these sensors on assets that are moving less than nearly 10 RPM. There's no way in hell. Even portable data collection is going to challenge you Exactly how long would you have to stand there to get the revolutions? This is a complete A minute. If it's doing 10 RPM at least Well exactly.
Speaker 1:You're standing there for a minute and don't get me wrong, I remember when I was at my old company. When, don't get me wrong, I remember when I was at my old company, we, when we had, we were testing these cookers. We were testing these cookers about four minutes on one reading, a long time. Waveform, to be fair, ultrasound is probably a better method of technique and measuring. Now, when we do that as well, in terms of just measuring for them, long periods of time. But again, this is another kind of real concern in the industry that you know they're trying to blanket approach and they're not even looking at you know even the speed. I mean, that's just basics. I mean it's scary.
Speaker 1:Do you know what I mean? That we're coming across that nowadays, because this is something that we're also concerned with, because you know how many times have we seen and we've gone to side and they've got, you know, little sensors and they're not even in the right position. In the right position, do you know what I mean? Like, let alone having the frequency range to pick up the defect. And this is where there's a clear misunderstanding of a lot of data companies selling sensors that have no reliability experience whatsoever or mechanical experience to understand how things can fail. Do you know what I mean? And it is worrying for us when we see this, because how many companies have already invested a lot of money in this and gone down a bit of a road and bit of a route with it, almost have to come back to square one to then do the re-evaluation process, because we wouldn't suggest you put all of these out until we've done the proper studies to understand why you need them and where you need them?
Speaker 2:yeah, because a wireless sensor, a continuous monitoring system like the mvx or the wireless sensor like the sparrow or any other on the market, doesn't make you detect failure modes any better. No, it doesn't. No, it doesn't make. It doesn't mean. Oh, if I put a wireless sensor and I can now detect this and I couldn't before. Portable data collection the only difference between these two sensors is that they can collect data more regularly 100% and also we have to consider.
Speaker 1:so I'll give you a really good, perfect example here with our kind of wireless sensor or online sensor solutions that are great, by the way and it covers a wide range, and this allows us as well to select the right sensor, depending on the machine as well. So we obviously use the Sparrow. Okay, still a very good sensor, right Five kilohertz frequency range, isn't it?
Speaker 2:Yeah, so the Sparrow is a MEMS-based sensor, so any sensor on the market that is of kind of a smaller form factor excuse me generally is going to be a MEMS-based sensor.
Speaker 1:What does that?
Speaker 2:mean, just so the guys understand.
Speaker 1:Basically throughout all of time. Piezoelectric sensing was always a way to do vibration analysis. Mem sensors more of a chipboard that has kind of more electronical chips in there doesn't have the piezoelectric element that usually is able to pick up the higher frequency ranges. But technology is improving, guys, and again, the more that we actually advance in this tech they're better these sensors are getting. We've tested the frequency ranges on these sensors. Very good, they're very good. Clean data as well, very clean. And again, I was always I was a bit of a skeptic at the start.
Speaker 1:When we say clean data as well, we just mean the data's not got a lot of noise in it yeah, it looks nice, you can see the peaks and, and you know, you can get the separation in terms of you know, like frequency ranges and frequency detection. But yeah, you know, that sensor very good. Obviously cheaper as well, because obviously you know to manufacture piezoelectric sensors is not cheap. Do you know what I mean? It costs money and it still can give you free axes of you know readings as well and temperature and temperature as well. So very handy metrics and measurements to know. So that's kind of the cheaper option. Then you've got the Eagle. It's still a wireless sensor option, okay, but I would say the Eagle's more designed to be able to replicate the handheld vibration measurements that we take but give you more of it if that makes sense, but you don't with the piezoelectric sensor.
Speaker 2:The way to think about the difference between the Eagle and the Sparrow, other than price point, is that frequency range basically means that the higher the frequency range, the most simplest way of thinking about it high frequency range generally, we can detect the defect earlier. Yeah, that's the easiest. It's not that simple but it's the easiest way to think about it. Because when a bearing no-transcript vibration that is going to start in the high frequency and the spectrum and as the defect develops and gets further down, we're going to start seeing that increase in amplitude, lower down the spectrum 100.
Speaker 1:So your piezoelectric, we can be a lot earlier with its detection, and but when we say earlier we're talking like generally months yeah, because, yeah, and if you understand that bearing defect where it comes from, what we're actually trying to see is the harmonic content of that, which will be at higher frequencies, yeah, okay. So when we see the harmonic content right, we're looking at higher frequency ranges to first identify that is available. So obviously, the lower the frequency range, the sensor goes down. You're not going to pick up the harmonic content of the defects right until it obviously it appears within its frequency range. So again, the Eagle great for looking at poor lubrication and stuff like that, because it can measure up to. Is it 20 kilohertz? It's exactly the same as our Falcons, which is incredible.
Speaker 1:Imagine having a portable data collector but taking data whenever you need it to. The idea of that is, though, you don't take probably as much as you would on the Sparrow, because you don't want to drain the battery, because it's a piezoelectric. So you're really looking at one or two readings per day. But then think about that, right, say, if you're taking a reading per month on that asset, right, that's 12 readings per year. If you start taking a reading a day, well, you're getting 30 readings per month now. So now you can now see the scalability of even taking one or two readings per day. It really gives you that understanding what's going on a lot sooner, if that makes sense as well, and that's the idea.
Speaker 2:Like both these systems sparrow, eagle, mvx all they're doing is allowing us to capture more data than what we would do traditionally with handheld equipment. Yeah, barrow, we can. You know, depending on how quickly you want to drain the battery, you might set it to one measurement a day. You might do something similar for the eagle mvx. We're looking at the machine every 80 milliseconds.
Speaker 1:It's as close to continuous as it is a wired system that's wired powered and piezoelectric as well and as well well, can integrate with a lot of different systems and grab information and display it as well A very unique system. But again, this system is really I'll be honest, is looking for different process parameters, really tricky assets to monitor. So whenever we go, do a handheld monitoring on the agitator, for example, well, one month it might be making product B. That is this viscous different load, different speed. It might range through speeds in just even making that one product.
Speaker 2:Last night I was talking with Aquam about an example on a CNC machine we're now using and what we can now start to do is we can take the information from the CNC machine into the MVX so it knows what tool it's loaded it's using Amazing process again.
Speaker 1:So when you're then trying to do handheld VA on real high process driven equipment, I'll be honest you are. It's a very difficult task, because if you can't trend effectively, you can't understand A what's good when it should be running and you can't deviate from that for a different kind of measurement.
Speaker 2:The MVX is used so much on wind turbines because a portable data or a vibration analysis worst nightmare is when speed changes, load changes, complex gearboxes. Wind turbines have planetary gearboxes.
Speaker 1:Yeah, and that's the thing you know. When we're doing vibration analysis, everybody knows it's so dependent on what speed that we're measuring, because if we don't know the shaft speed, well, how do we know the fault frequencies that exist from that? And if you get a constant machine that's constantly ramping up and down, you're never, ever going to be able to test that machine at the same conditions each time. So the idea is, you know, if you had a tacho and a shaft, you can tell the MVX okay, when it hits this speed and this condition. I want you to take data then. So then when we do look at our data we can say, ah, we've got a stable condition to trend against.
Speaker 1:But again, the majority of equipment, 80% of equipment, does not run that way on site, usually a high percentage of equipment or rotating machinery and quote me if I'm wrong. Do you know what I mean? I mean your plant might be completely different. You might have loads of that process stuff. You know what I mean, which is okay because we can still monitor it. It's just going to be able to be a bit more difficult, but a lot of the machines that we're looking at will run Fairly standardly. Up to a certain point it will run like that and it will run at the same conditions similar loads. Don't get me wrong. Loads may be a little bit different, but generally the same every time it runs, which means we have stable conditions for it to run, which means that handheld data collection actually is very effective. But then we now have to define the frequency of when we come To give you an idea of how much cheaper handheld vibration analysis is, because it is the reason why it's cheap is because of how it takes to collect.
Speaker 1:The data is very quickly, so within a day. If you've got assets that are running predominantly all the time OK, and you're not going to have this you're going to have a percentage of assets that are difficult to collect, but we've got solutions for that as well. This is another why of why we can use these other systems, because there's gaps in the data sometimes that we do need to make sure we've got, but generally we can collect up to 100 assets a day if it is running and it's consistently going. Think of that 100 assets that you've got in that one day. Do you know what I mean? That you've got that?
Speaker 2:And how many sensors you'd have to put on if you were going to-.
Speaker 1:So let's do a case study, right, real quick one. We'll just talk it out. We'll talk it out here, right? So you get 100 assets. Okay, in general, most assets will have a mower that runs or drives something. Okay, so it might be a mower that drives a fan, a mower that drives a pump. So if we average four points per asset, which I think is a good average, okay, that is 400 points. If you're talking about 100 assets, ok, 400 points, ok, right, is what you're going to have to duplicate to be able to get the same amount of data when you're going to be able to get from a handheld vibration program.
Speaker 2:OK say, for example do a little bit of quick math. Even if you simplified it and said the sensor cost would have to be cheaper than £5 a sensor to make it worth it.
Speaker 1:It would, yeah. So say, for example, right, just example, £350 for a MEMS-based sensor. We'll say £400, right, it's a good average. Okay, between most of the competitors On the market On the market, right. If we times that by 400, because that's 400 points that we need to monitor, that is going to be £160,000 on the off. We're not even talking analysis now, just in hardware. That is just the hardware. And that's a lower cost sensor. Okay, to be able to install on your factory. Okay, if, for example, say, va will do an average rate I'm not, you know, giving away rates anyway.
Speaker 1:Say, for example, a day rate is £1 pounds, right, picking out the sky, okay, within that thousand pounds, 160 000 pounds, right, divide that by the day rate, that's 160 days. Vibration analysis, just for the hardware. Divide that by 12. Divide that by 12., 13. So it's 13 years. 13 years just to be able to cover for the hardware. This is why, guys, it is not a wise strategy to go out and try to put a cent on absolutely every asset. It's not viable when you don't need to. Does that make sense? And I'll say that with my chest, because there's so many companies out there right now. They're just trying to blanket approach things. So 4.0, yeah, let's be the digital. Don't get me wrong. We've got incredible. We've got incredible systems. Aqua m have created some incredible systems. Do you know? I mean amazing systems, but remember we've got to apply the right technique for the right asset in the most cost effective way.
Speaker 2:I think that's going to then to understand really about the frequency aspect.
Speaker 1:Yeah, and this is where I'm going to talk about the frequency, because obviously this now does look into the reliability element and we are experts within this. So now we'll talk about OK, say, for example, we've made our point. Okay, it's obviously a lot cheaper to do handheld VA and when we do it it's highly effective because we're using high-frequency measurements and we're doing some observations. How do we define the frequency of how that is done on a standard machine?
Speaker 2:So we've spoke about already how Sparrow Eagle and MVX, basically the main benefits of these. Other than the Eagle and the MVX, they're also a piece of electric. These systems allow us to collect data more frequently and allow us to collect data without an actual person being there, which we'll come on to in a bit why that's quite useful, yes, and so the idea of having more frequency is that when we understand our failure modes, the whole reason is standard in the industry. Data collection generally is every month usually. That is the general status quo.
Speaker 2:Yeah, sometimes, on some machines, if the criticality isn't quite as high as others, you might be able to get away with a bi-monthly reading, but that is the minimum, because anything less than that, you're really struggling with your trends. High risk as well, high risk. So the whole point of setting frequency, though, is typically now you're not always going to have this reliability data to hand, so you have to use people with experience and expertise, and that's where the industry has set this monthly measurement, because, through data collecting over many years, we've established that, on most failure modes, a month should be sufficient, but it's determined generally from the P2F curve. So your point of potential failure. So if you're monitoring a bearing defect and you've got a machine and maybe it's a bit specialized and you understand that from the point of a potential defect being initiated to the point of it actually failing, if you've got that information or we can maybe look at ways of understanding it no-transcript then we should be collecting the data at half that interval.
Speaker 1:So that's where you get the monthly data intervals as well. And generally I'd say you know, that's a great, great point to make as well Will, because when we are choosing that frequency strategy, a lot of people will just say, oh, we'll do a day a month because that's what the standard is. This is where we also need to consider some assets on site that may be a little bit faster in operation. I'll give you an example okay, if you've got a two-pole mower that's running a grinder, for example, or a screw compressor or a screw, compressor gas compressor.
Speaker 1:Generally, these will run at high-speed intervals. Okay, this is where our monthly readers, you now may consider saying, right, we need to now take more data on these assets, and what you could say is, and what probably still would be cheaper is sets. And what you could say is, and what probably still would be cheaper is by just having another visit in between, because, again, yeah, it's going to reduce that time, okay, as long as the operating conditions are not deviating. All the rest of it, or then you could consider, so that could be a little bit overkill to send the engineer out. Then I would then consider the few assets that are running higher speed that are highly critical to the business. That's when you start to consider some online system. You then have also you've got.
Speaker 2:You can look at and go, okay, overkill, sending an engineer every month, maybe okay, even if in the p2f, even if we look at, let's pick this gas compressor and we say, well, you know, from the point of potential failure we might be looking at a week or two, right, you're going to want to measure it maybe every week or something like that. But you've also got with an mvx system where you're measuring it every 80 milliseconds. You might say, oh, that's overkill. But you've also got that safety factor.
Speaker 1:You've got that reassurance, not the peace of mind as well. And that peace of mind sometimes and that's the thing a lot of the time is what a lot of people say as well, when we're doing kind of vibration analysis oh well, it's been good for ages. Why do we need to do it? Because you're paying for the fact to know that it's okay as well. That's good information to have. But this is the other thing as well, and this is the thing for us with our strategy. When we look at it right, when we come to site, we want to ascertain where can we get a mixture of all of these things? So handheld VA, when we come to site and we get the good observational stuff as well, then we can start to look at some of the assets as well. And again, let's look at the reasons of why you'd want to put online system on and what would be our criteria for selecting the right. Maybe we should do that.
Speaker 2:Yeah, let's do it for each one. So MVX continuous data monitoring. Let's look it for each one, so MVX continuous data monitoring.
Speaker 1:Let's look at it. This is the hierarchy. So we're looking at the Creme de la Creme online system. Now that looks it's a powered system. It allows us not to worry about battery changes. In terms of sensors, it still uses PSU electric sensors as well. We can use dual output sensors as well. We can look at, obviously, temperature sensors as well, and obviously everything can go into this, and there's no limitations with this system.
Speaker 2:Maybe we could do a bit of a backwards and forwards of our ideas, why we'd put an MVX for each one. So I'll do the first one. Yeah, you go. Okay. Varying speed or load is why I would consider putting potentially an MVX on, because we can input the speed and load into the MVX, so it collects data all the time.
Speaker 1:Perfect me health and safety. So if you've got an asset like a fan, a big fan, where there is a, is this high likelihood there's going to be? Yeah, there could be a change very quickly in fault, say, for example, imbalance, right. Say, for example, you've got a fan and generally it may run and it may not. It might ramp up and ramp down we've had that of a few fans as well. But but usually say, if there's a shocking balance or something comes off the impeller and it goes out through the roof in terms of vibration, we want to know now, we don't want to wait until next month because the failure, then the risk of failure of that asset, multiplies dramatically. So you've got also look at them failure modes and what could be the real risky failure modes that could cause that asset to completely spiral out of control.
Speaker 2:So health and safety a real good one, yeah, for me as well uh, another one, okay, uh, machines where the failure modes can happen very rapidly, yeah, so things where, um, something can happen and it can deteriorate very quickly and we may need the system to be able to identify that a lot earlier. So things like that.
Speaker 1:Yeah, definitely so. Yeah, mvx is looking at real high process parameters.
Speaker 1:It's live, it's live another example is like an agitator. So these would be these agitators. For us have been a right pain to be able to monitor. Okay, because every time we go to site it might be running a different product, different speed, different load, doing all sorts of crazy process parameters. An NVX system will be able to set that up, to be able to take data at the same, similar loads, same speed, so we can get a good, reliable trend, so we can still analyse this data.
Speaker 2:Machines that start and stop very rapidly, yeah.
Speaker 1:It's so difficult to you know you're putting your sensor, we're doing it now you know it's going to go here, then here, then that yeah, over to there.
Speaker 2:So we need to be able to trigger that data very quickly.
Speaker 1:So you are looking at very high critical machines yeah, that could have post or complex machines or complex machinery that are difficult to be able to gear through, handheld and even wireless. Because you could say, well, why don't you just put a wireless system on there? But let's, for example, say you put a Sparrow on it. Well, a Sparrow window, in terms of data collection time, is what? Probably 0.4 seconds. That's going to play.
Speaker 1:You have to think the best way and as well, with the Sparrow, you can't obviously trigger that reading to start. It will just every X amount of time it will take reading, it will just wake up. So you might be missing the reading every single time trying to wake. The likelihood of getting that information or data is very low, let's be honest. So let's go to eagle now. Okay, so eagle is really what you're looking at. Is your data collector being able to take data a lot more regularly? Okay, so with this particular application, I would not suggest it for high process parameter changes as well, because it's not designed to be able to do that. Because, remember, the Eagle really is designed I'll be honest to take a reading a day really designed, I'll be honest, to take a reading a day if you, if you clock it to take more readings, that is, you're going to vastly drain the battery because, remember, it is a wireless sensor and we're taking high resolution piezoelectric data, so you can't trigger that to take 10 reads a day.
Speaker 2:You are going to drain your battery you know, I mean even once a day is going to be somewhat extenuous on the system, depending on all the measurements that you've got set up on the eagle. All depends on what you're taking as well is where it comes in. As to really where the eagle is kind of pitched out and designed for, which is one of the best applications for eagle is a hazardous environment that you can't exactly because the one thing about the eagle it is this is a zone naught.
Speaker 1:Yeah, it's a zone naught, ex. So perfect opportunity where you probably don't want a guy to go into because of the health and safety element of it, but you're still able to take data every single day. Do you know what I mean? You know and get that decent trend over time as well. So then it then takes a risk away from an engineer going into there, into that area or hazardous area. But still, I mean again, usually if assets are in these hazardous areas, they're going to be fairly critical, aren't they? To health, safety and production and environment probably as well.
Speaker 2:Yeah, you may not want your engineer going there but that Eagle sensor is going to give you, especially for when you maybe have complex machines that are still continuous in their running speed, so they're still, their load is the same. Or you need that earlier indication, so say you maybe have got, maybe it is a gas compressor, but the load doesn't change, the speed doesn't change, it's always running in this operation you want to be. That's a very critical asset to the plant.
Speaker 1:It might be in an area that's really hard to get to or poor access, and that would be an ideal yeah, we, we are using again within our industry I'm looking at the biogas industry, we are looking at mvx systems and eagles to be able to counteract both, because obviously they're hazardous environments. Depending on the strategy, it's either one of them or a handheld data collection with an EX data collector as well. So sparrows obviously. Again, these are the beautiful little sensors as well. They are because what we can do as well now is change a battery very quickly, which means now we can take a little bit more data than what we are looking at with. You know the eagles, ok, so that means we can get a better picture of what happens within the day with that. Obviously, the more you take data, the more it's going to drain the battery and the quicker you're going to have to change it. But because that now has become a lot cheaper and a lot easier to do now, the option is now on the table. But because that now has become a lot cheaper and a lot easier to do now, that option is now on the table, but as well, what it does allow us to do is able to be able to again, it's a lot cheaper sensor to look at the lesser critical assets still important, but as well, gives a good idea what's happening with them as well. So a good application use case for me would be if we're doing monthly VA and we're finding that you've got you know you look at your data and all the assets we're testing. We find that E102, for example, we haven't been able to collect data on that in the last three months and that's quite concerning because it's fairly important to the plant. But every time we've come to site, unfortunately it's not been running, it's not going, it's not running. So we find it quite difficult to get a steady trend with our VA, which then creates a longer gap of data, which then creates higher risk for that asset as well, because we don't know what the condition of that asset is.
Speaker 1:This is where I see sparrows coming into action. Okay, you don't necessarily want to put high piezoelectric sensors on these assets, because it could just be a gearbox or small motor gearbox and the cost of that is going to still be too much, right? And not only that, you still should only take one reading per day with that. It's really designed for assets that run all the time. What you can do with the Sparrow is take more data within the day. You might get a bit of missed data, but it's going to be able to detect when it is running and you're going to be able to get that information to fill in the trends on them.
Speaker 1:Assets that are really difficult because of process right Again access Okay, you might have an actual fan up, I don't know, up out the way, hard to get to Our sight out of mind. Do you know what I mean? Sparrows, for these assets are inaccessible. Now I think I'm going to supersede the fixed sensor model. So back in the day we're just well. Well, I fixed sensors to it to go to our data collector machine. But now the cost of these sensors is fairly is a lot cheaper. Why not put a sensor on so we can actually start to get more data on it as well? Seems to me to make more sense in that, in that action as well. But remember, even if you put fixed sensors on this, it's still relying on it running. So you still you know if it, if it is a application, well, you might not get data on it anyway.
Speaker 2:And Sparrow as well is a great option for plants that are looking to go down a digitalisation approach, but they need to understand it's OK to say it, to be in a position to say, hey, we want to digitalise the plant, but it's important to understand the why you want to digitalise the plant, because digitalising the plant isn't going to make you detect the defect any necessarily more effectively.
Speaker 1:What it's going to allow you to do, though, it's going to allow you to be a lot more predictable when the defect becomes yeah, and you say digitalize the plant, that doesn't mean that handheld is not digitalizing your process, because nest vision allows us to see all of the data that we collect in one place, which is digitalizing your process. It doesn't matter how we take that data, whether it comes from a wireless sensor or we're doing it. It still allows you to see your health conditions of all of your assets, just in different ways. So when we talk about site digitalization, that's still a form of it. It doesn't have to be constantly monitored by a sensor to be digitalized. Do you know what I mean? We're still driving data into the same place, but we're just picking the right frequencies, the right strategies, the right techniques to be able to mitigate the failure modes in the most effective way, in the most cost effective way, because everybody knows, even that 100 asset model that we just went over there a lot of sites have more than 100 assets. So you have to now remember the cost of being able to put a sensor on everything without really knowing the reason. Why is not the right strategy, guys? The right strategy is looking at the whole thing, then having a good mixture of different things, depending as well on the failure modes of your equipment. So if we do identify the failure modes are difficult to detect.
Speaker 1:Well, we'll put a few mvx systems on on really high critical. We can't afford to lose, we can't afford to lose them assets, and they might be moving in different ways. So we have to make sure we monitor that way. Then it's having a look at the next stage down, saying, right, we've got a few hard to access machines, highly, highly important environments, environments as well, so we don't want anyone going up there either. So, great, perfect. Well, let's put some eagles on these, we've got covered there. We tick all the boxes there. Then we might go a step down and start to look at some other things that are a little bit inaccessible, hard to get to, not as critical and maybe not in a toxic environment, but then we still can then deploy that Sparrow sensor kind of option.
Speaker 2:It's on the site still, but not at the point where it's important enough that you kind of want a bit more than a monthly data. You kind of want a bit of visualisation of what's going on, but it's not to the point that you want to put you know a 10,000, 15,000 NVX system on or, you know, with the Eagle sensors, because the Eagle sensors are a lot more, because you've got a lot more high tech.
Speaker 1:Yeah, it's a high tech sensor, really it's a great sensor, but you're going to have to pay for what you get for.
Speaker 1:But then, when looking below that, if you haven't got that criteria for any, why then have an engineer on site?
Speaker 1:It's accessible, it's easy to get to the failure modes.
Speaker 1:Don't allow it to fail within that period that you've said with the PNF curve, handheld VA is going to be extremely vital and it's going to be a very important strategy still to be able to get the bulk of the data on a high number of assets in and around your plant and no matter how that we gather the information, it all can go to the same portal with pictures and also can give you an idea of the lifeline of all the events that we found within that asset along the way.
Speaker 1:So, over time, when you talk about digitalizing the process with your asset health condition, this is exactly what we mean by it, right? This is the way to package it all up to be able to understand how do we now see and how do I know my asset health across the whole plant and then we can look at it in terms of that way, and I think for me that is the, the strategy that we're moving forward with, because a is the most cost effective. B you're utilizing the best technology, but only when you need it. But you've got data to back up why you're doing it and it will only ultimately, you know, have it.
Speaker 2:Putting a sparrows sensor on is fantastic. It's such a great tool on the majority of machines where nothing's too complex, where we can get that measurement, you know, once a day if we really want to. And it may be that the future, as the plants digitalize further and further, that more of these sensors get deployed.
Speaker 1:But it will only happen as the price model becomes more and more and more competitive with yeah, and also will only happen as the price model becomes more and more and more competitive and also, it will only happen when everybody on the plant has done the reliability standards and understood how everything can fail and have put a good plan maintenance system in place to counteract the fact that we're not there Exactly and that, for me, is going to take a hell of a lot longer than the latter, because I think technology is getting cheaper and I do think is going to take a hell of a lot longer than the latter because I think technology is getting cheaper and I do think it is going to, you know, Technology will get cheaper before the reliability is embedded Because the culture remember technology.
Speaker 1:We've seen this mate through our whole journey within this business growth. Right, the technology moves at a rapid pace, but the culture and the understanding and the reason that we're doing is not moving. This is why we're here, this is why we do this podcast all the time, to raise the awareness of why we're doing it. Let's not just do it, let's understand the reasons why let's have a discussion, let's have a chat about it, let's understand why let's have a look at the whole purpose. Remember, we're only looking at just one small part of the pie here. We're looking at a vibration analysis strategy. We're not even talking about thermal image and talk about oil sampling, talking about all these other things that as well will be identified in the formica stuff. Remember, this is just us coming in and almost looking at that one side. There's a lot more we can and will help with off the back of it. 100. We've just gone deep into one subject, if that makes sense as well. So remember, there's a wider scope than all of just vibration analysis.
Speaker 1:You know you talk about, you know cm strategy or a condition-based maintenance strategy. Yeah, it's great to understand there's a defect in the, you know, in a gearbox, but that's too late. It doesn't improve late, you're not improving it a lot. You're detecting the issue, but I can guarantee if you're seeing where within a gearbox, well, the gearbox is warm, it's too late. We need to understand the org and that's a whole nother story for. But again, that taps then into proactive maintenance, which is a completely different thing. Right now we're detecting issues. There's the detect model that we're looking at now.
Speaker 1:Right, the improved bit is a completely different culture conversation. Do you? You know what I mean? And I don't know if the industry is quite ready for that yet. Do you know what I mean? Because when we're talking about proactive maintenance, well, if you're not even detecting the issues that happen and you're in the reactive state, then we're not ready for proactive maintenance quite yet. We've got to get to us to a position where we're not reacting. So remember, if we're reacting, we're not proactive. So the first aim is to get out of reactivity first, so then we can focus on the next stage. But I think that opens up a new line for the next podcast to talk about proactive maintenance. What are the tasks that will prevent failure? Do you know what I mean?
Speaker 2:and improve liability, the tasks that are often done whilst we're collecting portable data collection.
Speaker 1:Yeah, definitely Might be another really cool podcast to go over, because I think we've covered today really nicely with kind of vibration, analysis and strategy right. And if anyone wants to have a chat about us, because we can come in Demo in any of the sites, doesn't matter how big your site is, how small your site is right, we've got demo kits on all three of them systems as well, to show you what the data looks like when it comes in. We've got trial versions of Nest as well Nest Vision, so you can see how it looks like for your assets to be in that portal and, as well. We could even do a little bit free. We could take three or four assets with a bit of with handheld, put one on the Sparrow, put one on the EVX, show you what it looks like so you can see the whole strategy. But, more importantly, we'll look at all of them assets and we'll devise a strategy of what is the best strategy to move forward with, and we can do that quite quickly.
Speaker 1:So, guys, thank you for tuning in. Really great discussion today Got me amped up and ready to go. I'm actually got a meeting now to go talk about some Sparrow. I'm ready, I'm ready to go. So, guys, thank you for tuning in and yeah, we, yeah, we'll see you on the next one.
