Manpreet Singh with Nanoprecise

SUMMARY KEYWORDS

Nanoprecise, energy-centered maintenance, machine doctor, IoT, predictive maintenance, energy consumption, rotary equipment, AI algorithms, maintenance professionals, asset management, efficiency, renewable projects, CMMS integration, sensor data, talent crunch.

00:00

Scott, welcome to the industrial talk podcast with Scott Mackenzie. Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots and let's go

00:21

all right once again. Welcome to industrial talk, the number one industrial related podcast in the universe that celebrates industry professionals all around the world because you're bold, brave, you dare greatly, you innovate, you collaborate. You are changing lives each and every day. That's why we celebrate you on this platform, and we are broadcasting on site. We're on the floor right here. They're still doing work. In the background, I'm looking at all of the carpet that still needs to be laid down. It's happening. But we're here SMRP. And you know what I forgot to do? You know what I forgot to what year is this? This is SMRP, 30 something. I'll mumble through it. 33rd 33rd I guess. Guess 33rd See, you're right. Your little question sketchy on that one too, and I forgot to do it all right, we're here 33rd SMRP, and it is a collection of incredible professionals, problem solvers, and we're right on the middle of the floor. We get to see it all, and we get to talk to everybody. It's all great. In the hot seat. We have a net gentleman by the name, name of Manpreet Singh Nanoprecise is the company we're going to be talking about that we're going to be talking about energy centered maintenance, which I like, I'm all in. So let's get cracking. Yeah, there. I'm worn out. How's that?

01:41

You don't look like you're one

01:43

I am. You know, as it is, we haven't even started the show. Are you excited that is this the first time Nanoprecise has been here?

01:51

No, no, we've been here for a while. Have we? Yes, come on.

01:55

How come you? How come you? Okay, I've been here a number of times, and I don't remember what you guys have been doing, but I like it all right. Before we get into the conversation, let's give us a little background on who you are and why you're such a great professional and how long you've been going on. I mean, just give us, give us a little background.

02:18

So I am an engineer by profession. I started my industry in my career in hardcore manufacturing industry. Yeah, we were building pressure vessels, heat exchangers for oil and gas, fertilizer, pressure, I mean, all of process plants. And that's that's where I started my journey. Then I went on to do my business education. I moved to a company which was setting up renewable projects. So I did that for about seven, eight years. Then I got a chance to move into IoT. And technology was something that I was always very close to. And when I saw IoT solving a lot of our problems on the shop floor, maintaining those plants, I felt really connected. And it it sort of became like a side hobby, where I would build multiple projects on my own, and as I moved on from there into different industries, did my own startup and post that I also moved into health tech, but that IoT pace always stayed with me. And then I moved to Canada,

03:29

and we're about to Canada. I'm in Toronto. Yeah, you want to know a little secret, yes, all my family lives in Hamilton and Dundas. Wow. I'm a my cat

03:40

fan. Next time you're there, let's, let's,

03:43

let's meet up. Here it is. Maybe I'm all in. Okay, continue. Yeah. So

03:47

04:05

head honcho. He's the man right there sitting right across from me, delivering some sage insights. There you go. Now, let me ask you this real quick, because I am a generation nerd. I love anything dealing with generation. What type of renewables were you involved in?

04:27

I was in solar. Solar.

04:31

Yeah, that my pedigree. I used to negotiate power purchase agreements for Southern California, Edison and and in all you know, renewable forms,

04:43

you and I were doing the same job. I was also negotiating PPS, setting up those plants.

04:49

There you go. Look at that. All roads lead together. That's such a small world. All right, let's talk a little bit about just just give us a little background. On Nanoprecise, what? What's what you do. And then we're going to talk about energy centered maintenance. Thank you, because I got the RCM, but I had to slip in energy in there. Yes.

05:16

So Nanoprecise is a startup in this field, and we have built technology to work with machines. Our product is called machine doctor. It essentially senses different parameters of the machine, and then, just like a doctor diagnoses What's wrong, it prescribes, what do you need to do to make the machine work right? And that's that's essentially what we do with our sensors and our analytics. But over the years, we realized that it's not just that the machines fail and we are able to predict those failures, but before they fail, there's a lag, and during that lag, what happens is these machines are running inefficiently, like if your bearing is off, it's going to consume more energy, to run that shaft or run that fan, that pump right which we all know, like if our cars, we don't maintain them, they start to consume more gas. And that's essentially where the topic of energy centered comes into RCM or PDM, because you are essentially trying to link a mechanical issue in the machine with its energy consumption

06:29

back up, just because I love utilities too as well. Yeah, your solution is not just specific. Or if it is, it's not a big is it specific to assets that are in movement? Because in the utilities, there are assets that you just can't they're they're moving, but you can't see Yes. Do you have that solution as well? Do you look at it that way too as well?

06:58

No. So we are primarily targeting the rotary equipment, like you mentioned, which you can see rotating or moving, and that's that's where we play. And if you ask me, more than 50% of global electricity consumption is is through those rotating machines or the motor driven machinery.

07:17

So do you have, yeah, do you have any sort of insights into how much energy is wasted through, you know, inefficiencies in failure of the degradation of those assets? Because I'll just be honest here, we have to be more efficient. We have to be more efficient in our power generation, in our utilities, we have to be more efficient because we're just not building more infrastructure anytime soon. Yeah.

07:49

So in terms of energy consumption, if, if a motor is running brand new smooth versus when it's running with a fault, you can see energy and consumption go up from anywhere for for a very minor lubrication issue to go up by 5% of its rated capacity. But if there's a major fault, it can be as high as 25 30% sometimes even 40% extra. That's a lot. That's

08:16

a lot. And when you start to look at it from a from a macro perspective, all of the energy that is just being wasted through that. So it's, is it safe to say that that your solution on the asset spinning identifies, let's say, a potential problem. It's, you're, you're compressing the time to correction, absolutely, that's that's where the that's where the nugget of gold, exact

08:43

size, exactly. So two things, if I look at the life cycle cost of a motor driven machine, 75% of that life cycle cost is electricity. And if you can reduce that consumption by 10% over the lifetime of the machine. It's a huge saving. Second thing, like you mentioned, what we're trying to do here is, traditionally, predictive maintenance was about predicting failure so that you could stretch out your maintenance schedules to align with your plant shutdown or the next availability of the material. But what we're telling people is you didn't have this additional information when you were doing predictive maintenance, what is your energy footprint looking like? And we're not saying that you have to be sustainable for the sake of sustainability, it's actual dollars that you're spending on energy, and that's why we're giving another dimension to you, so that you get another input to your decision, making that okay, I save 100 bucks by delaying this maintenance to my plant shutdown, but I spent 200 bucks just on energy during that time. Is it? Is it a good decision?

09:52

See, that's that's interesting, yeah, because we also talk about sweating that asset, extending it out, pushing it. As far as out you can before you perform maintenance. But if you bring into that equation, the the the energy component, then it creates a different decision tree of saying, no, if I extended a little and then and now, we're upside down. Let's change it here.

10:16

Yes, that is the paradigm shift that we were bringing into the industry.

10:22

Yeah, see that that makes sense? Okay, you can't, you can't just have a conversation about, hey, that's fantastic, which we have been, because it is fantastic. Got it? It's more in line with, Okay, now that I'm collecting this data, and it's data, and I'm pulling data, and everything's fine. Where are we? Where are we putting

10:41

this data? It sits on our platform. So, so you have a platform, yes, so most of our users get access to the sensors as well as the data that these sensors are generating over time, and we are able to also correlate the mechanical health of the machine with the energy consumption profile, and that gives you an actionable insight. I can tell, I mean, I can put energy meters on my machine, but machine consumption can go up because of the process or my own requirements that I'm running a different process. But what we differentiate is if it is changing because the process has changed, or whether it has changed because an underlying mechanical issue has come up and you need to address something

11:24

with the machine. Is that in the secret sauce? Yes, yeah. I was just gonna say because that that makes sense. If I'm running up, if I'm wanting to run in a liquid process, and I'm running a more viscous, let's say, product through then, of course, it's going to have a different energy consumption profile than if I'm learning water or something that is, it just is, yes, okay? And I shout out to CMS radio, okay, they're down there. And I have to have this conversation, because now I'm collecting data. I'm running through the process. I have my assets. My assets reside within my mice, my platform, my CMS platform, and I want to be able to attach it. So I don't want an additional system. I want to go into my CMS and and be able to access all of that great information in there, because I'm too lazy to go over to this other system. Yes,

12:18

and that's that's something most of our users have asked for, yeah. And the way we have designed our systems, we have kept it extremely flexible, so we have API's available anyone who wants to integrate to our system. We are already doing integrations with maintenance, maintenance. We're also here, yep, and that's, that's, that's the way forward for us, like we build those connectors to every CMMS out there, and whichever system you're using, we you'll be able to consume whatever insights we're giving you in the way you want, in your day to day workflow. We don't, we don't want you to change,

12:56

yeah, your workflow process, and say, Okay, now jump out here and go over to this system and be able to access this information on this particular asset, and then jump back into another system. Yeah, that's like, no, that that that dog doesn't hunt really well.

13:12

And for us, if the systems can talk to each other, we also get feedback, like, if you did a maintenance on the machine, that's very useful input for us to train our algorithms further.

13:22

Yes, bi directional. So we, we've, we've touched upon it. So we've got a great, great platform identifying all of the potentials that can happen with that asset. It's energy centric, centered maintenance, which is great now, bi directional. So once again, I'm, I'm taking, I'm taking information here, going into CMMS, then I'm sending it back over here, because it's bi directional, right?

13:54

Yes, it is, and that's where the value comes from, right? Over so our models are trained to look for patterns, and over years, they will, they will look at how these patterns are evolving. But not all patterns will lead to some problem with the machine. Some are just patterns which are created because of a transitionary condition or something that happened, so that feedback helps train those algorithms better, and sometimes we find a lot of value because actual end users who are on the shop floor, if they tell us that this pattern is not because of an underlying health condition, then we can go back and train our algorithms to ignore similar patterns in Future, or have something with a with a better confidence, telling that, okay, this is, this is a consistent pattern that we have seen in the past, where you did this maintenance section and we are bringing same context now to this current

14:55

15:43

usually deal with the sensing parameters of the machine, not really the process parameters. Some of the things that you said might be changing the process parameters, but we are. We are creating a repository at the back end for ourselves, for all the machines that we monitor so that tomorrow, if we get the same machine with the same nameplate or same similar Yeah, footprint, then we know what kind of bearing was used, what kind of running speeds Does, does that machine typically run at? And that helps us detect patterns for newer machines far more easily than what we would do for it to train over time.

16:25

So that makes sense. That makes that's the right, right solution, because now you're, you're leveraging the data that you're you're collecting just because you've got clients, and you're pulling that information, and you're, you're providing that level of insights. Where do you see, I mean, everybody's talking about AI, everybody, including me, I guess. And, and, where do you see Nanoprecise solutions? Where do you see it going?

16:53

So we've already had AI in our algorithms for a long time now, even before llms became popular in a household name, we were using AI algorithms, yes. But now, with llms being there, we are taking it to the next level. So we are now able to consume all the historical context of the machines, all the physical feedback or manual feedback that we had obtained, any pictures of the machines that we have, all of that can now go into a bunch of agents, like there are five or six agents that we have developed. Each agent does its own piece of analysis. One just looks at the patterns, generates the diagnosis, moves it to the next agent, who looks at whether I've seen something similar on other machines, on this machine in the past, what happened that time? So it's gathering all of this context, bringing it together. And now the insight that it generates becomes conversational. Like I tell you, you're at the shop, you're if you're hearing something, let me know I'm seeing this pattern. Are you seeing something similar? So it becomes very conversational with the people on the shop floor. And what we are looking at is augmenting their skills. You are in the industry for a long time, and you're seeing that talent crunch has become really, really a big problem for the industry, where a lot of maintenance professionals are not available. What we're trying to do here is help the maintenance professionals make better decisions, because they now have access to all the data from various machines, and they can converse with the platform.

18:35

Man, pretty you bring up a very interesting component, and that's, that's something that I've been on, and that is, is, how do we become more efficient? But, but we have, apparently, limited human resources. They're not just people. Are not just knocking down the door to become maintenance professionals, which is sad, because it's a great profession, and I think we need to do a better job at, you know, inspiring the next generation. That's just me. That's going to be my theme for this week there. But, but if we can programmatically, through technology, be able to make our limited resources more efficient in how they conduct business. Yeah, that makes sense. But, but again, the whole conversation about what we were talking about is about efficiency, gaining greater efficiency, which we need absolutely in the world of power,

19:33

absolutely, and you can't have a maintenance professional Manning every machine. And these machines are consuming energy. They are. Every machine is running with some fault at certain stage. It's not visible, but it's there. So you need, you need to improve the efficiency of these people to make sure they are able to take the right decisions. And that's what we are after. I like it.

19:58

I like it a lot. I. How do they get a hold of you? They're saying, I want to know more about this.

20:03

They can, they can look me up on our website, or they can write to me at and sing at Nanoprecise.com where I'm mostly available and I do it. Are you in LinkedIn? Yes, I'm on LinkedIn. Look me up at, don't worry about it.

20:19

I'll have the link out there. I don't worry about it. Well, you were just absolutely spectacular, man, please. Thank you, Scott. I enjoyed the pleasure. Oh, wow, no, it's, it's an honor on my side. I'll tell you that much. All right, we're gonna have all the contact information for Manpreet out on industrial talk. You know that it's gonna be his LinkedIn stat card, which I'm sure is just chock full of just incredible. Well, what a professional. I loved it. Nanoprecise is the company. They're here. We're broadcasting on site, SMRP, and again, we get people like Manpreet right there, wanting to help you succeed, become more efficient, right here. All right, we're going to wrap it up on the other side. Stay tuned. We will be right back.

21:07

You're listening to the industrial talk Podcast Network.

21:17