Mike Kahn of Elemental Machines

Welcome to the macro fed engineering podcast a weekly show about all things engineering, DIY projects, manufacturing, industry news and IoT data sensing, where your host electrical engineers Parker, Dolman and Steven Craig, this is episode 327. Our guest this week is Mike Cohn.

Mike Cohn has been working in the high tech industry for over 30 years. The bulk of this time, he has been involved in the architecture and design of complex electronics, and embedded systems across a wide variety of segments, including general purpose computing, digital communications, digital video, medical devices and life sciences instrumentation. Mike is currently the vice president of device engineering at elemental machines, a startup that has created a service that allows lab ops professionals to monitor their lab equipment through a combination of IoT devices and a suite of cloud applications.

Thank you, Mike, for coming on our podcast to talk about what we're gonna talk about.

There's a variety of machines.

And like enterprise consumer, Nolan not known that could suit No, not consumer electronics and enterprise electronics. Which is that's I'm super excited to hear about.

Sure. Thank you for having me. I looking forward to the conversation.

So I think we'll first start out though, Mike, is how did you get started in all of this?

It goes pretty far back actually, I I first had my first exposure to a computer when I was in the seventh grade, and we got an apple two. So that shows you how old I am, I guess we had an apple two with the the high school that I went to, or my middle school rather. And I kind of fell in love with technology at that point. And I was pretty sure I wanted to do something with computers when I went to college. And then from my career, I think originally, I thought I was going to do some sort of software development. And my next door neighbor, who is about six years older than me actually majored in computer engineering in college. And it was really interesting to me. And so I decided to pursue that as a college major. And I found that I had a real interest in the hardware side of things, in addition to software. So one of the great things about the Computer Engineering major, particularly back in the late 80s, is that you could go into hardware or software depending on on what area interested you. So I came out of school. And I guess, I started working my way through I think when I was in college, what I wanted to do was to design computers, you know, general purpose, computers that ran user code, didn't really understand a whole lot about what the embedded industry was all about, didn't realize that that's probably significantly larger than then general purpose computing. And so I started out working for a company that made Unix servers. That company no longer exists. But I did eventually find my way into communications. And you know, as I look back on my career, I think every change that I've made during the course of the career wasn't necessarily planned. But looking back through the history of things that I've done, it's always been a logical step. So I went from computer design. When I got into the communications industry, I started designing really the Compute Engine for switches. And then I started picking up that whole communications piece. And then when I moved into digital video, that was really, I worked on a video on demand streaming server startup that did a video on demand streaming server. And that was really a switch. And so it was that communications expertise, moving data through this switch that gave me the exposure to digital video, when I moved into medical medical devices. That was a product that we moved digital video around the hospital operating room. And so there was this natural progression. And then eventually, when I stopped doing actual product development, moved into medical devices, excuse me into the life sciences business. A lot of that was around understanding quality management systems, how to how to take complex electronics and scale them and build robust electronics. And then that was what got me into the life sciences business. I spent six and a half years working for a company that makes instrumentation that's used by the life sciences industry. 50% of that company's businesses is pharmaceutical. And that's where I saw this need for lab operations and lab managers to really start to get controlled or instruments because if you look at a typical pharmaceutical lab, whether it be in QA QC or on the research and development side, they have this wide array of equipment from multiple different vendors. And it was nice to see what elemental machines was doing to bring all of this information about how that those instruments were working into one single location so that people could then monitor their instruments without having a million different bespoke implementations of that. So yeah, I guess that's sort of my career story there. And in a nutshell,

well, the end there seems to segue really great into what elemental machines is, can you give us a rundown on it?

Sure. So, you know, we provide a service to lab ops professionals. And, you know, you think of lab ops in the same way that you think of DevOps in the computer world. DevOps people started out as just, you know, their SIS admins, right, and they had a very narrow focus. But over the years, they took on a broader and broader role within within fortune 500 sort of environment and, and what we do for lab Ops is the similar kind of thing where originally I think people thought of them as lab managers. And their role is becoming broader and broader and really are concerned with the whole operation of a lab. And so we provide a service for those lab ops professionals that allow them to monitor the equipment that's in their lab, and then make intelligent decisions about how to spend money on either capital or operational types of reasons. And the way we provide that service is through a set of IoT devices that are either dedicated sensors or data movers. And we connect those IoT devices to the cloud. And we have an application in the cloud that allows the users to then see what's going on with the instruments in the lab, we also have this analytics piece that is taking some of that data that we have and giving them insights into how their their instruments are running.

Just perhaps from my ignorance, can you can you give me a rundown on what a data mover is, I can guess, but just as yours.

Sure, I'd say it's a it's a I guess it's a term that we coined in our device team. So a lot of these instruments that you find in a typical farmer or biopharma lab, they, they might have a bunch of sensors embedded inside of them for actually running the instrument. Take an incubator, for instance, where the chemist is growing the culture or the biochemist is growing some sort of a culture, they actually have to do a closed loop control on temperature, and maybe humidity or co2 levels. And so of course, in order to do that level of control, you need to have sensors embedded in the instrument. Well, they provide that same data on an interface on the back panel of the instrument. And we can just tap into it with either a digital interface for Rs 232 45. Sometimes it's Ethernet. And we package that information up and then move it up to the cloud for our application. In some instruments actually have an analog interface four to 20 milliamp that you find in a typical industrial application. And, again, it's just in our case, we're taking the the information off of that four to 20 milliamp interface, packaging it up and sending it up to the cloud for our application to then report to the customer.

So it sounds like those interfaces are quite varied in how they get exposed to the world on the back panel, as you said, because we were talking a couple months ago, Steve and I about electronics, lab equipments, and usually more modern stuff, at least, is there's a standard for it. It doesn't sound like there is a lab standard, so to speak,

there. Yeah, I mean, the lab instrument industry is huge. And so and I think it's largely lots of companies operating very independently. So there is no, there's no historical standard, a lot of these interfaces, it's just straight serial interface. So that's actually pretty easy for us to replicate over and over again. And we actually have a team within elemental that we call our integrations team, that they go and they pull down the datasheet for the the instrument. And then they create a configuration file that gets loaded into our data mover device that is really just a mapping function, right? It says, Here's the address, and here's the data that we're getting. And then it tells it, how to package it to go into the cloud. There are some other standards that are used. We have some instruments that actually use Modbus. So we have a Modbus implementation. But there's also some industry specific standards. There's one that's called OPC XML, which is used by a lot of analytical instruments. We don't have a lot of interactions with that, but there are some instruments that we have done an implementation of that, that OPC XML that's run over Ethernet. So it is it is pretty varied. But we have you know, we have a dedicated software developer who actually is responsible for doing that higher layer protocol that's running on either the serial interface or the Ethernet.

So you aggregate all this data and package it up nicely and provide it in a in a convenient way to display to the customer. Right?

Yeah, that's, that's absolutely true. Our cloud application is a dashboard. And really the primary, one of the original applications was just alerting, right. So if you think of one of our dedicated sensors as a temperature sensor, and you know, I think during the pandemic, the entire country has learned about minus 80 and minus 100. Freezers that they've been storing the vaccines and right, well, it turns out that these same freezers are used in farm these research labs for storing their cultures. And as you can imagine, if one of those freezers goes in the middle of the night, the it costs a lot of money, either in the sample gets lost in the sample may be worth a lot of money, or there was an a lot of time that was invested in creating those samples. And so we have a sensor that it's a small device that has a long thermocouple. So the sensor sits on the outside of the freezer, the thermocouple, then sneaks into the into the freezer. And then every 15 seconds, we send a temperature reading up to our cloud application. And the user has the ability to go into the their dashboard and set an alert. And it says if the temperature drifts outside of this region for a period of time, then alert me and they can get the alert with either an SMS text message or voice call, we have an app, the mobile app, so they can get a push notification, or an email and allows them to act on that alert almost immediately.

Very cool. Do you also have any controls that gets sent down to the devices? Or is this one way communication?

It's pretty much a one way communication, we don't actually take control the instruments, we're just monitoring how they're they're behaving?

Yeah, that reminds me a lot of it's like a live version of SCADA, I guess, which is like industrial control, where it's like, oh, I think they have two way communication. But mostly, it's one way. Very, that's interesting, because I'd never thought of like, thinking of like, a lab or anything like that. Set up that way.

Yeah, I think historically lead the lead managers haven't really thought of it either.

Well, you have to have your interns or whatever, nothing, right?

That's right. Yeah, I think the way I view it is, you know, there's someone there's a clipboard on the front of each freezer. And, and, you know, I check the temperature at this time, and it was 37 or minus 80. And this time, it was minus 80.2. And, and it just really takes a lot of that out of the equation. And I think it's, again, one of the things that the pandemic taught us is, how do you do all this stuff remotely, they didn't want to send people into the lab to do that monitoring. So if we can do it remotely through a dashboard and monitor an entire building at one time, it really, it creates a healthier environment, and also allows them to be a lot more efficient.

Kind of reminds me, probably two years ago, I signed up for a service that allowed me to connect some IoT devices and just have a dashboard up. And it was very much a hobbyist grade thing. But seems like overall, it's the same general concept.

That's right. And then on top of that, we actually have a data science piece of this as well. So we have a dedicated analytics team. And they're actually trying to draw some interesting conclusions about some of that data as it comes across. So for instance, with that temperature sensor, on the minus 80 freezers, we can actually detect when the freezer door was opened. And so we can even we can even do some predictive things like your temperature went out of range. And maybe it's because the door was opened 15 times yesterday. And so it helps, it helps them to manage that aspect of it too.

And do you allow the customers to write their own perhaps alerts or, or, I guess some of that predictive information back to them?

Well, all of the predictive analytics is done is done by us. And we're, you know, we're in the process of rolling a lot of that out right now and gaining new insights into the data that we're collecting.

So you're the vice president of device engineering, what is your primary focus at elemental machines?

So as I was discussing, we have the device side where we do the sensors and the data movers in our in our IoT gateway. And then there's the cloud application and so the development of the of all of the hardware is in my organization, the hardware itself, the embedded software and the embedded firmware, as well as system level verification and then and then the design aspect, but the primary focus of my organization is on that development of the devices. And then the actual cloud application is, is in another organization. But all that said, you know, we're one big happy family within engineering at elemental, because it really is a system that we're delivering. And so when you're delivering a system, you actually have to make sure that everything goes together very nicely. So while we don't actually take a system engineering approach to this, we do work very closely across both cloud and devices, so that we can provide the best possible solution for our customers,

you don't just design something and tuck it over the fence. Exactly,

exactly. I mean, at a minimum, you've got the interface that you have to make sure it's clean. But But that, you know, there's some interesting aspects to this. As a networking guy, I've always, I always think, in terms of how network switches are built. And I talked about data planes and control planes. And, and I try to make analogies to things that I've worked on since then. And so going from the sensors up to our cloud application, that's what I think of as the data plane. But there's a whole control path, that that needs to be considered here as well. And some of the control path is entirely controlled by elemental. And so we use third party services for some of that. But some of the control path is, is really up to the customer. And it has to do with how we configure these devices, you know, when when you're dealing with IoT devices, they, at least one device needs to be on Wi Fi, so you can get out to the cloud. And so convenient and Wi Fi is a bear to deal with. Because it's it's so widely varied that. So we have to do a lot of this configuration through our dashboard as well. And that's where there's really a lot of integration and collaboration between the device team and the dashboard team. Because the dashboard team, they know how to do the web applications, but they don't really have a good feel for how the devices themselves work. And so we have to work very, very closely together to put together a seamless sort of experience for our customers in the dashboard, even though it's managing something that happens in the device.

So my next question is about more about the hardware, dig a little deeper into that sure is, how are they powered? You are mentioning Wi Fi? Do they piggyback off the devices themselves? Are they just like a wall wart plug or?

Yeah, so the depends on which device you're talking about. So in some of our devices, so our gateway, for instance, is Wallwork, right, that's that gateway itself is powered off of, you know, your mains voltage, and it's got a pretty beefy embedded Linux kind of computer in it. It also has battery backup, and a cellular backup as well. So they can communicate through a variety of means the preferred means is either Wi Fi or hardwired Ethernet. But you can imagine that if you're monitoring your freezer, you really want to monitor when the power goes out. And if the power goes out, you've lost Wi Fi. And you've also lost power. So that's why we have the battery and the cellular backup in there. So that's powered off of mains voltage, our dedicated sensors are actually all battery operated, and they are not connected over Wi Fi. They're wirelessly connected to the gateway. And then and that's why we're able to to run off of a battery for those

that are both sorry, like, the wireless connection to the gateway or using like 900 megahertz or what frequency is that? It's 2.4 gigahertz. Okay, so just not. So it's like Bluetooth. Okay, okay, cool. Yes.

So yeah, so that's the sensors. And then our dedicated data Movers today are Wi Fi. And you know, the the reason that those are Wi Fi, we take the one that's the the analog, the analog one, it really it doesn't need to be necessarily. Because it really doesn't require a whole lot of compute. It doesn't require it but but the approach that we've taken to a lot of our hardware development is to do something that is kind of lean startup ish, in that we try to get a product into the hands of our customers as quickly as possible. And then get as much validated feedback as we can before we harden into something maybe battery operated. So that element, what we call the element C is running off of an off the shelf computing platform. And we designed a custom daughterboard for doing the it's really just an add on it. And then the next step is to harden that into something battery operated because really all it's doing is an A to D and then pushing it over, over the Bluetooth back to our gateway. So today that's means means voltage powered, but in the future that will be battery operated one on the, on the digital one to trick your question because doing a digital interface, depending on the digital interface may require a bit more power. And the other problem is with or the other challenge with digital interfaces is that, you know, you take something like this OPC XML that I was mentioning, that's connected over Ethernet and the instrument might want a DHCP server. So, we're not doing that in some microcontroller. And so, you know, figuring out exactly how to get that into a battery powered application, that's actually something that's going to be coming down the road.

I just I have to go back to you using the word hardened on that. I'm absolutely going to steal that now. For all of my prototyping at work. Instead of admitting that my prototypes have issues. I'll just say that they're a little soft, and we need to. Glad I could be of service. Yeah, that's fantastic.

Yeah, I've actually never heard that term being used, like hardened. Like the example I use, like locked in or something like that. I like that term, though. I'm not a big fan of Stephen softened. But I do like the hardened term.

So every customer is a unique client. And with all unique applications, right, so are you handling each client individually in that way? Do you have like application engineers that work with them?

Yeah. So I mean, when you say unique, I think you're referring to maybe the kinds of instruments that they have correct? Yeah. Yeah. I mean, as it turns out, all there's a there our best sellers, I guess. So, you know, we've done integrations for the most popular incubators first. And then. And then as we get new customers, and we learn about new integrations that they need, like I said, we do have an integrations team that handles this, and they take it on a case by case basis. So I wouldn't say that it's a custom for every customer. You know, I think that we can hit the bulk of their instruments right out of the chute. And in and then if you think about, you know, the, if they're really just going for temperature monitoring, that's something we don't have to do an integration for at all. They just, they buy our sensor, and they connect it to their freezer, and within minutes, they're seeing results on our dashboard.

Sure, I guess I was meaning no two labs are going to be exactly identical.

That's absolutely true. Because they're all doing different things. Yeah.

So in addition to temperature, sensing, what other things Are you sensing.

So we also have another sensor that does what we call ambient sensing. And so that has four different sensors inside of it, it's got a temperature sensor, a pressure sensor, a light sensor, and humidity sensor. And so that might be used for monitoring conditions in the lab, because those kinds of conditions can actually impact experiments. Or they might some customers actually will put that inside of an incubator that doesn't have sensors of its own. And where they didn't want to buy any off the shelf sensors. So we might use it in situations like that. So that's, that's our second oldest product behind the the minus 80 temperature sensor. And then we have a recent product that was released just last year, which we call the element U and U stands for utilization in this case. And in that one, we can actually strap this to a power cord in a non invasive way and and measure it's got four Hall effect sensors on it. And we can measure the magnetic field and infer something about the power that's going through that through that power cord. So we're not getting actual current draw up, but we can get relative current draw and make decisions around how that instruments being used. Is it on? Is it off? Is it idle? Is it actively being used right now. And that's actually where a lot of our analytics are coming into play. And we think that it's a relatively new product. Like I said, we just launched it last fall. And we think there's a lot of opportunity in labs to be able to take advantage of a product like that.

It's a easy way to sense that something is happening, right?

That's right. And you know, there are the advantages to doing it this way with the Hall effect sensors as opposed to your typical current monitoring is that you don't have to rip open the power cord to get around one of the legs of your of your mains voltage. We can actually get it by just going around the power chord. It's a completely non invasive way to do this and as it turns out, some of the instrument vendors have this particularly of more complex instruments As they you know, if you do something to their power cord, or if you stick something in line between the power cord and the wall voltage, it voids the warranty. So that's one of the reasons that we've chosen to go this route, it makes the installation significantly easier, and it allows them to, to really be confident that nothing's going to happen to the instrument because we'd cut into a power cord or something like that.

And that's on the on the AC power cord, right? That's correct. Yeah, I'm gonna have to, I actually have like a Wikipedia page about Hall Effect current sensing around AC current. And I'm like, I never even heard of that before. So yeah, that was that's a problem with like, just regular coil. For current sensing is the alternating current, you just can't, it looks like zero, right?

That's right. And but it all has to do with wearing the twist, you're actually sensing it? Because they're those are twisted. And so the way we've got the product architected, we can, we can.

Oh, no, I'm just saying the normal way a current sensing is Oh, yeah. Well, that's right. So if you're passing, that's actually how GFCI works. Oh, yeah. Yeah, yeah, sure. It's when GFCI is, if it detects current on AC, both wires, it goes, Hey, there's leakage somewhere tripping. I'm gonna have to I'm gonna have to read up. Unless you can just discuss that right now. Like, that's actually super fascinating. It's like, how did the Hall effect sensors work on that?

Well, because of the fact that you've got this twist going on. And I'm not really an expert in this, we actually have an enormous amount of feasibility analysis that we did on it. But it's, there's a, there's a, it doesn't always cancel it out, depending on where you're measuring the twist. Okay. And also, of course, the, the safety ground is going to shield it in some places. So we can detect a magnetic field somewhere. And so again, we're only looking for relative changes in that field, as opposed to what the field absolutely is that we can determine, you know, through our analytics, exactly what's going on in there. Interesting.

So what kind of industries or are you involved with with the elemental machines?

So right now our focus is, you know, we say life sciences. It's pharma, biopharma biotech. You know, we're located in Cambridge, Massachusetts, which is a biotech hotbed. So we have certainly got a lot of customers in that area. But but we really are geographically diverse. We have customers all over the country, but it's, it's really mostly the the biotech biopharma kind of space, we started out with smaller companies, and we are expanding into larger and larger enterprises now.

Steven, do you want to move on to the next topic

there? Yeah, let's let's go for it.

Okay, so what's different about elemental machines? And I guess most guests that we bring on is y'all pretty much focused on enterprise customers, instead of consumer customers? And I've never had to deal with working with enterprise customers in terms of developing a product. So I'm, I would be very interested in how is that different from a consumer standpoint, the development of product versus enterprise? Developing a product, I should say?

Sure. So it's a really good question. And I think it also depends on what your definition of enterprise is. You know, throughout my career, I have worked for a number of different organizations that have had quality management systems, whether it be ISO 9000, or ISO 1345, or to 9000. And I think that's one of the things that that maybe differentiates us, or all these organizations from something that's more consumer minded is these certifications that we have to get in, you know, having a pretty sophisticated quality management system that really prescribes the taking of the user requirements, translating them into a set of product requirements. And then creating all the documentation to support that you've met those product requirements and user requirements and then having a verification and perhaps even a validation on the back end, and demonstrating that you've met all these requirements. So I think it's really the main thing for me is the level of rigor in your overall process going from, you know, from from business development all the way until you actually deliver a product that has some some value for the customer. And in fact, elemental we just achieved ISO 9001 certification, this earlier this year. which was a huge effort for us. But you know, it's funny because I think people, especially if they've never been really exposed to a quality management system think it's, it's a lot of overhead. It's, it's, the engineers hate it, because because it's paperwork, but at the end of the day, it's a lot of best practices. And you hope that the people that are doing consumer products are following the same kinds of best practices, maybe they're not, maybe they're not following it, you know, with with documentation, like, like we have at a place like elemental, but it is really best practices. And you know, one of the great things about ISO 9000, in my experience is that they don't tell you what your process should be the edits, at its basic level, it says, have a process and make sure you're following it. And then take a risk based approach to working your way through it. So you know that the first place I was exposed to this kind of thing was at my first job. And ISO 9000 was brand spankin. New at that time, there were very few US companies that were getting certified. And I remember they built a process that was like three feet thick. It was, it was incredibly burdensome. And I think it's because nobody really knew what they were doing at the time. And it was a process that we couldn't, we couldn't hope to pass. Yeah, exactly. And we failed our first audit as a result. And, and it's a learning experience. And then the second company I went to, same thing, ISO 9000, but their process was like one page, and it was very easy to follow. Now Now what we've done here at elemental actually is built a process that is largely based on how engineers work. And, you know, our VP of quality has documented the process that we're following, and says, what, let's just keep following this process, it seems to be cranking out good products, let's have all of the artifacts that show that we're following it. And then we can, our customers can be assured that, that they're going to be getting a high quality product, because we are certified to the ISO 9001 standard. And actually sounded kind of like an ad for ISO 9000. But I do think that having a good quality management system is really a key. A key piece of being delivering to enterprise level customers.

I've I've said on this podcast before is like if ISO 9000 is important to you, it's not just the make sure they if you're the same manufacturer, your manufacturer has ISO 9000, it's you also have to make check to make sure what that process is, because as you were saying earlier is like it can be one page long, their process, and they're just following it all the time. But it might not cover everything that needs to cover. That's right.

It is however, we've mentioned it multiple times and not to boil it down to too low of a level, it does give a little bit of warm and fuzzy feelings when you see that whoever you're working with is ISO certified, because you know, they had to go through something to get that. And it and it's not particularly the easiest thing.

And it shows a level of commitment. Right that to your point, they went through this because they were committed to making sure that their customers knew that they're delivering a quality product.

Absolutely. Right. Yeah. And once again, not to boil it down. But grabbing grabbing a point of data from a thermal couple and displaying it somewhere isn't necessarily like the most difficult thing. But doing it with all of these, I guess regulations in a way and backups and quality is what separates it.

Yeah. And rigor with Rick Rule. Exactly. And what's interesting is that while we're not actually regulated, our customers are regulated. So they're, they're expecting a certain level of quality, because at the end of the day, a lot of them they have to they're subjected to regulatory audits, and it's it's audits that are based upon the data that we're providing. So having that ISO 9000 certification gives them the assurance that that we're going to be providing a high quality level of data,

their freezers at point two degrees

well, even to that power cable, you're talking about where you they could void the warranty with that they're in a way they're self regulating the their own warranty, and you're, you know, supplying solutions for that. That's right. Whereas, I guess you could just go get a current clamp and slap it on there and say, good enough, let's go. Right,

right. But But again, with a current clamp, they would need to cut that power cord open

exactly as I'm talking about, right. Yeah, that's, that's a disaster waiting to happen. That It's

particularly an eye, a lot of our, a lot of our customers, the people that we actually interact with these lab ops managers, they're not electrical engineers, they're, they're not computer people. In fact, they usually have to interface with their IT departments. They're guys that know how to buy lab instruments and keep their lab operating. And, you know, there's a whole lot of other factors that they can consider. So they rely on us for doing that sort of electrical interaction and, and really make it problem free for them.

Actually, out of curiosity, going, going back, do you have any kind of phone apps for being able to display data or I guess it's just, it's available in the cloud through a browser, right?

We actually do have applications as well. And the primary reason for that is they can receive push notifications, I guess that's the primary reason they can receive push notifications in the event of alerts. But yeah, they can view how the various instruments are, you know, the current conditions on those instruments. So someone you know who's out hiking, or something they might want to see, you know, they get an alert, right, and this is the main thing is they get an alert. And all that the alert says is this, this freezer has drifted out of range. And so they might go into the app to then take a little bit closer look at it, and

then discover that someone has been opening the door 20 times, it's not the same as you could put it out on your your personal fridge and tell if your kid is just not in front of the fridge? Yeah,

I do. Actually, I have one in my house that for a while was connected to my personal freezer, and it was kind of depressing seeing it get open so many times.

You know, one thing I'm curious about is on the enterprise side of things, when it comes to manufacturing and designing for enterprise, do you see that there is any kind of a push for? This is an odd question. But like the attractiveness of products? Or is enterprise Okay, with here's a black box with a thing that you stick on the side of your, your bridge? And you're good to go?

That's a That's a good question. You know, we do certainly care about industrial design we have, we have an industrial designer on our team. And, you know, part of it is because you want to make these easy to install on. So that's a big piece of the industrial design. I don't know that, that anybody cares how something actually looks in a lab might be easier

to sell to people who don't know anything about what it is, but they're the people who signed the check.

We definitely, and I think the other piece of of having a decent industrial design is that you just want it to look like a quality piece of equipment, right? I think we could, you know, we could 3d print these things if we wanted, and but they wouldn't look very good. And I think it would defeat some of the quality that we're trying to go for. So the aesthetics itself, I don't think the people working in the lab care all that much, but but there is definitely an aesthetic angle to the design of the elements themselves.

Yeah, I guess what, what's what's going through my mind now is, it should look like it belongs there. And it should look like it isn't going to break.

That's, and I think those are those are two keys, certainly the it isn't going to break is a really, really important one. Because we do want them to really think only about the application. And the element is just sitting on their instrument doing what it needs to do. And they should have confidence that it's going to continue to do what it needs to do. For sure.

On the enterprise stuff is in my mind, at least, this is how I because we were talking about like, what do you mean between consumer and enterprise? And for me, it's when you're designing this is just me, when I designed if I was designing something for enterprise, usually, to me that is a company or is a group is coming to you to request stuff, or a certain product, and you're going to be selling to like one specific group. Whereas a consumer product might be you're doing market analysis on your end to figure out, like, what it might be the next big thing, this is just like, and then and then designing that. We're sounds like you're kind of like right in the middle of those two things. So how like, how often do you have to actually do I think we touched a little bit on this, but how much do you have to actually do like custom hardware software?

There's, we never do custom hardware. It's all you know, quite generic and actually that's going to be important as we start to looking at new verticals that we want to expand into the custom software really is is more on this, this element, the data movers element the in the element, yeah, how's it really

chasing with whatever random machine that they want to get data from? Right,

exactly. And you know, a lot of those machines, they just provide a serial interface. And then we just map to what's coming off of it. So the integration is actually we've written our software in a way that it's just a configuration file that gets loaded into the element, and then they can pull it pull it out, but you know, every now and then an opportunity will come along, where there's there's some new interface that's running over serial over Ethernet, and will look at the opportunity and say, Is this really something that we want to make into a product, and then we can do some custom software development to do that.

And, you know, speaking of different verticals, with the first time I visited this, your website, and I saw lab Data Sensing, it made me think back to my first job, where I was designing safety equipment for vibration, sensing and things of that sort. And in our lab was mainly an electrical engineering lab. But we had plenty of equipment that, you know, we would leave on overnight, and we temperature cycle things or, you know, we had spurious events or something like that, that we were trying to capture on an RMA. So we'd have, you know, our vibration table running for three, four hours trying to find some kind of error in the system. And I could see having a system like this be great to not so I don't have to just sit in my chair next to the thing, wait for something to happen. So moving into something that's, you know, perhaps, like engineering lab also seems like a like a, like, it seems like a good parallel.

Yeah, I think there's lots of instrument types that we can certainly connect to, and I'll put you in touch with our business development manager. So yeah, I mean, we certainly, like I said, you know, the data movers in particular, they're, they're fairly generic. And so we can we can move data at a lot of a lot of different instrument types. As we move forward and start to look at new verticals. Yeah, we're, you know, we've built a solution here, the IoT devices, the cloud application and the analytics in the middle. And that whole solution can apply to lots of different applications and, and, you know, we provide best in class of all three of those. We're the only ones that are kind of putting all three of those together to create really the give the ability to allow these lab managers to problem solve and really optimize the costs associated with running their labs.

I can just imagine it now the lab them. I'm working with now if I had all my multimeters and all my oscilloscope and everything, where I could just leave work and leave my product on and monitor it overnight. That'd be awesome.

That that's a great vision. I'm gonna write that down

we, you don't even

Yeah, yeah, I think. I think yeah, I think that's it for me.

You have anything else, Mike?

No, I think it's actually been a really enjoyable conversation. And, you know, I appreciate the fact that you invited me on, talk to you guys.

So Mike, where can people find out more about you or talk to you? Same thing with elemental machines.

So I'm on LinkedIn, you can find me on LinkedIn, that's probably my most active social media spot. But elemental machines, we're all over social media. Elemental underscore AIOs is our Twitter account, but we're on LinkedIn, as well. So or go to our website, www dot elemental machines.com. There's plenty of information there as well.

Well, again, thank you so much, Mike, for coming on our podcast and discussing all these topics with us.

Thank you for having me.

So that was the McWrap engineering podcast we're your hosts Parker Dolman

and Steven Gregg. Later everyone take it easy

Thank you, yes, you our listener for downloading our podcasts if you have a cool idea, project or topic. Let Stephen and I know Tweet us at Mac fab at Longhorn engineer or at analog E and G or email us at podcast at Mac fab.com Also check out our Slack channel. You can find it macro lab.com/slack