Voiceover: You're listening to Augmented Ops, where manufacturing meets innovation. We highlight the transformative ideas and technologies shaping the front lines of operations, helping you stay ahead of the curve in the rapidly evolving world of industrial tech. Here's your host, Natan Linder, CEO and co founder of Tulip.

Natan: Today on Augmented Ops, we're being joined by Matt Lowe. He's the co founder and CEO behind ZeroKey. I met Matt for the first time climbing up a ladder, but you'll have to listen to the episode to get the full story. More specifically, I think we're all familiar with traditional location tracking systems like RFID or GPS.

That's what ZeroKey have been working on and develop a completely different way. They call quantum RTLS, Real Time Location System. Not gonna get too technical, but it's involving ultrasound, echolocation, wireless, and all sorts of intriguing tech, but it really works to change how operations work, allowing people to do material tracking, error prevention, quality, and all sorts of other good stuff.

So let's take a serious technical look and see what it's all about.

Matt: Hi Matt. 

Natan: Howdy. Welcome to Augmented Ops. I'm super excited to have you join us today. We're going to talk all about spatial intelligence in manufacturing. Can you believe it? 

Matt: Well, thanks for having me. And, uh, excited to talk a little bit about, uh, spatial intelligence. Yeah. And what that means for the factory and beyond.

Yeah. 

Natan: So Matt is a classic, uh, back of the napkin type of idea person. Started in running a company called ZeroKey that makes a real time location system. And we're going to learn about, and we kind of figured this episode, half of it is going to be nerding out on tech and understanding like, what does it actually do for customers in the real world where we try and go deep on like how operations get augmented by all this technology stuff.

So Matt, maybe we start simple. So for people who don't know what a real time location system, RTLS is, maybe explain and then let's talk about your journey founding Zerokey. Okay. 

Matt: I think we all use RTLS pretty much every day. So real time location system in its most basic form tells you where you are.

And so we all use GPS, whether we're calling an Uber or, you know, using maps and navigation, you're using a location system. Now, there's different types of systems and our system is focused on the indoor environments to enable digitization, digital ecosystems within factories. So it's more focused on the indoor space where things like GPS don't work.

Natan: Yeah, it's been kind of a holy grail of so many applications over the years. Very short anecdote of how we actually met in person for the first time. This is roughly two years ago, and this is how I know you should really pay attention to a CEO of a company and look at their technology when you're considering partnerships.

So maybe one lesson to take to heart here. Basically, the first time I met Matt in person, he was like Hanging on a ladder, installing their stuff in the Tulip experience center. And I was thinking like, Oh my God, you know, they're showing up 24 hours, like a crack team before the event goes live. And, you know, we kind of talked maybe 48 hours before, and the CEO is like on the ladder, installing their tech and working very closely to the engineers, making things work, and then.

Coming up with like kind of a perfect demo. So that was to me pretty inspiring. I don't know if you remember that this way, but, uh, that's what I remember. Yeah. From my 

Matt: perspective, it was interesting because, uh, you know, we were jumped on a quick call late, I think you're right, like 48 hours beforehand, and we were just talking about the different ways our two technologies could work together.

And it was just like doors opening on ways that like, now we can address this and address this and customers can utilize it in this many other ways. Yeah. But, you know, we had kind of a short notice to get things up and running. And the thing that blew me away and, you know, it's kind of a source of pride on one side is that, you know, within 24 hours really was the, uh, the lift to get, you know, these two technologies talking together and working together in such a seamless way where that demo could be pulled off.

And that's a testament to how well the technologies were built to work with other things, right. To be a part of that digital ecosystem from the ground up. So that, that was kind of cool. Yeah, I 

Natan: definitely think we have a great engineering teams on both companies that did amazing work. But I also think it's not just how well it was built, it's also the principles it was built with.

Right. Because if we were not building this thing open, it doesn't matter how great the tech team and you and I wanted, thought there were like great use case A, B and C and we can do whatever, you know, we can decide we want to do a show, but if it's not architected correctly, forget it. It's going to take two years to get stuff talking to each other.

Matt: Oh, exactly. And like, I've been a part of a lot of software projects where that's been exactly the case, the experience where, you know, we think we'll have something done inside of, you know, half a year and you learn, oh, it's not speaking the same language for one reason or another, and you're two years into an integration project versus it's virtually unheard of for technologies that are doing such advanced things that are like complicated behind the scenes, but providing kind of that, that frontline value to get those talking to each other and up and running in the span of 24 hours.

I think, you know, it shows exactly like you said, the design principles are really showing through and the investment that has been made to really make sure that the technology is ingrained with those design principles, you know, right from the first stage. 

Natan: This is a perfect segue because what we need to cover now is what is Quantum RTLS?

How did you get to it? What does it do differently? So we can talk about the cool application of this thing. 

Matt: So Quantum RTLS is the name of our product line and our technology for indoor location. The big difference, you know, and there's a whole bunch of little nuances, but the big difference is our accuracy.

So we will locate where something is in real time down to 1. 5 millimeters. In 3D space. So, you know, roughly in the order an eighth of an inch in full 3D space. So that's a pretty big game changer in the types of applications you could even use RTLS for. Because historically, like if you look at GPS or you look at Bluetooth or ultra wideband or some other, you know, indoor location technologies, these talk about position on the order of Meters or, you know, 3 feet and beyond.

And usually on a 2D basis, not a full 3D. So, as soon as you step into the world of 1. 5 millimeters in 3D space, you now have a very very accurate image of exactly what's happening in these different environments. And, and that means you can drive. Smarter solutions, better solutions that precisely react to what's going on as opposed to a more coarse position fix that it would not give you that level of detail, wouldn't give you that insight into, hey, this guy has used this tool in the wrong spot, or he skipped torquing the bolts on a jet engine that's going to end up on a plane, that sort of thing.

And so it really opens the door for things that haven't historically been possible. That's a 

Natan: great description, but we've been hearing about these types of systems and every time I would talk to people about their RFID, Bluetooth, Wi Fi, Ultra Wideband, all of them, you kind of get this like, yeah, it's working, but Kind of working, but is your stuff different?

If so, why is it different in that sense? And what do you think all those other technologies are actually good for? They're not all kind of like completely irrelevant now, right? So help people as an expert in this domain, understand the range of stuff out there and what's good for what. 

Matt: Yeah. So the big thing is, and this is really a testament to the market.

You know, people have always wanted to digitize these types of physical processes. And so the demands on things like Bluetooth and ultra wideband have always exceeded the technology's capabilities. So we've always wanted accurate position so that we can do tracking of precise manufacturing operations.

But the technology just wasn't there to do it. And so, in a lot of those cases where those customers are having friction and having trouble actually getting value out of, out of the technology, it's because there was a misalignment between what they wanted and what could be provided. Now, where we're different, and by no means am I saying we're the only technology for all of the different solutions, you know, I believe that there's the right tool for the right job, but when it comes to a lot of these high value applications, you need the accuracy.

Without the accuracy, you really don't have better information about what is happening in a production environment. And that's really the big difference where we're succeeding, where Others have failed previously. I mean, you can take one of our sensors, attach it to a tool, and watch in real time, or even replay it historically, the exact tool path that the operator took, and exactly what part of an engine or a widget he torqued the bolt on.

And now it just starts compounding the value that you can get from that. So you can start to do real time quality control, real time. Integration of your manufacturing processes. So it's just layers and layers of value get added on when you have that level of accuracy. So I totally 

Natan: see it, but all these technologies, you know, maybe one way to group them are really just pure radio based.

That's right. They're prone to noise, right? They are really affected by the environment in which they're installed. What would be the main pointers you would give to someone looking at this? Thinking about getting into RTLS for whatever application, you know, accuracy we discussed, but what else should people consider?

Matt: You know, I think we've all had the experience maybe driving in a downtown urban area and our GPS has us in the wrong location. Like it's bouncing around, uh, streetwise. Yep. And this is a problem called multipath. You basically have these radio signals and they're bouncing off of different things. And you know, it's a time of flight based solution.

Those bounces increase the time of flight. That's why you have a bad fix. When it comes to the indoor environment, the multipath problem is maybe 10 times worse. Because indoors, you've got machinery, equipment, whatever products you're working on, and they're all reflectors of radio signals. And so we're We differ a little bit using ultrasound.

We don't have the same degree of multi path difficulties. And there's a whole bunch of reasons for that. Without getting too technical, one of the real reasons for it is that ultrasonic signals disperse when they bounce off of things, as opposed to a RF signal reflects almost like a mirror, like light off of a mirror.

And so that makes it very difficult to solve the multi path problem in the radio sense. But in the ultrasonic sense, it actually makes it really easy. It actually helps you to not have all those bouncing and reflecting signals. That's maybe a simplification of the overall problem, but that's fundamentally one of the reasons ultrasound is better.

One of the other reasons, and this is a key part of how we can achieve our accuracy, is that when you use ultrasound, it's, you know, a sonic signal, just like any other sound that you hear. And it means it's traveling slower. You know, speed of light, roughly 3 million meters per second, speed of sound, 340 meters per second.

So you've got orders of magnitude difference in how fast that signal is traveling. Well, one of the advantages of a slower moving signal is that you can actually be much more precise in timing how long that signal took to get from point A to point B. There's some pretty Heavy math behind all of it, but it's really one of the big enablers of why we can get to that 1.

5 millimeters. And in the indoor environment, it's like a perfect solution. It doesn't have multi path, it's more accurate to measure, and it generally just works 

Natan: better. But what about noise in the factory? Doesn't it interfere with your 

Matt: stuff? Generally, no. It's really a kind of a couple different reasons why.

Our technology is an engineered system. So it's engineered to be resilient to noise. That means we leverage a lot of the things that you leverage in radio systems for like coded signaling so that, you know, you're not just picking up random noise. We use a very Narrow band within the ultrasonic spectrum that you don't have a lot of natural emitters of noise in that spectrum.

And then the other aspect is that ultrasound starts to attenuate faster than, uh, lower frequencies. So we, we've all heard, or, you know, probably had the experience of a, a neighbor who's like got their stereo cranked and you can hear the lower frequencies, the base coming through. Those signals go further and don't attenuate as fast.

The higher frequencies attenuate fast. Most of the signals that we hear in music, for example, start to cut off around 15 kilohertz. You know, when you talk about ultrasonic, you're talking well beyond human hearing. In our case, we use 50 kilohertz. So you can imagine the attenuation is even that much greater.

So even when there's You know, some sort of source of like a natural ultrasonic emitter that dies off very rapidly. And so all of those things kind of work in our favor in the indoor environment to where ultrasound is a pretty phenomenal technology for positioning in that space. Great. Before we move on, let me give you a good example.

And we have this problem sometimes even with customers that are in the space. It's not always clear the value of accuracy in positioning. But I think we can actually look to a couple examples historically where you can see the difference in improvement that it made. And like one of the clearest indicators, like the aviation industry, you know, until kind of the mid to late eighties, there was still like sextant navigation through spy glasses in the, in the planes.

Like that was the standard approach for it. And you could say in that, Particular case, you know, lots of planes landed, they made it to their destination safely, but they took rather roundabout paths as they continuously course corrected, you know, it was really inefficient in terms of the fuel burned. And it wasn't until GPS came in that all of a sudden things got way better.

Now you're flying direct, you're no longer off course, you're not worried about where you are, you're not getting inaccurate weather reports or think that you're in an area where you're avoiding weather. Meanwhile, you're, you're in a different position. Uh, so it really kind of transformed that industry.

And that all happened on the back of a rather tragic accident. It was a Korean Airlines flight back in like the early eighties where they didn't realize where they were. And so they had, unfortunately, on a flight to Seoul, flown off course, ended up getting into, at the time, Soviet airspace. The Soviets, given that their airspace was infringed upon, ended up shooting down the aircraft.

And that was what the impetus was for the U. S. government to greenlight GPS in airplanes. To avoid that situation of flying off course, and this is moving now from like an inaccurate system to an accurate system, and you get all these other benefits, and it's kind of the same in the industrial environment, the production environment, where, you know, we've got some good indicators, we've got all sorts of sensor systems, or you Maybe one off systems spun up to like handle specific problems, but we're still having to course correct on all these little things that come up in the manufacturing or supply chain process.

And we're doing so with very limited inaccurate information. So now if you can take that high resolution position information, know exactly, you know, where things are going wrong, when they're going wrong, and course correcting immediately when they do, as opposed to, you know. Down the line, when you've got a quality control step, you're saving time, you're saving money, you're decreasing yield loss.

I mean, the benefits just keep stacking up. Yeah, totally. 

Natan: And, um, we're going to tell three stories on how the integration between ZeroKey and Tulip solve real problem from customers. But before diving into that, maybe you can describe very quickly, what is that integration and what was that? for you and your team in general and how that helps implement those apps.

Yeah, 

Matt: so I'll tell the integration story first, because it's actually one of my favorite stories. And this came after we had built the components over that initial 24 hour period. And then we kind of refined them in the following weeks. We had this goal to build a PIC tracking system. You think of it like, Pick the light but on steroids.

This came from a customer use case. And what had happened was one of our marketing people ended up in the loop just in the back and forth with the Tulip team and ended up falling into being our Tulip expert within the company, which The marketing person. The marketing person. Typical startup fashion.

Natan: Yeah. Is that person an engineer? 

Matt: No, not an engineer. So that's the interesting thing about this story. You know, they came from marketing backgrounds and no engineering experience, no software experience, no experience really building systems. And they were able to basically take Tulip with the widgets and stuff that were built previously, take one of our starter kits off the shelf.

So these are off the shelf on both sides and within two hours had built the picking system. To do validation of the pick from bins in real time, basically from scratch. And that's, you know, two hours to go from a non digitized picking process that is typically found in every factory floor in the world to a 1.

5 millimeter 3D position controlled validation system, real time validation system that every factory in the world would be salivating over if they could do that. And that's how easy it was. Literally two hours and no special training, no engineering background, nothing. 

Natan: Awesome. And if people want to see that integration, it's on the library, we'll include some links and people can check it out and see what it does.

But let's jump into what people are actually building. So we have three quick stories. One around warehouses. The other is about the management of critical inventory in luxury goods with a cool application. It probably has one of the coolest names, you know, it's like called Double Verify. We'll talk about that.

And a story about right first time. So maybe we start 

Matt: with the warehouse. Yeah. So this is one of our most interesting use cases and I like it because it's so simple and the simplicity of it makes it easy for customers to implement, but it's driven by that really rich data, the position data. And so this is a customer, it's a typical warehouse.

50, 000 square foot warehouse. They've got, you know, maybe on the order of 20 warehouse staff and they're, you know, relatively high volume, like all day long, they're busy, orders in and out. They had this problem effectively that when goods were received into the warehouse, they didn't always know where they were, you know, despite the fact they had a binning system with barcodes and, you know, very typical, very standard process.

And so they were spending a lot of time looking for things. They're spending a lot of time not finding things, restocking. ERP. Do they have an ERP? They had an ERP. Yeah. They were like relatively standard as far as like, this wasn't a bad warehouse. It just, this was a warehouse that was, it was fairly typical of the size it was.

And they were having issues, you know, not finding inventory, like losing inventory, which was quite expensive for them. And they even had a case where they had picked the wrong part. On a very time sensitive order for a hospital that was shut down, uh, the wrong part went out to the hospital and as a result, the hospital was down for an additional, uh, I think it was like three or four days and you're talking rescheduled surgeries, all sorts of things like a real human impact to that mistake.

And so we built them a solution using Tulip in Zerokey and it was really simple. We weren't doing any fancy analytics or anything like that. All we did was every time a barcode scan. happened, we tagged the 3D location of the barcode scanner. And with that information, we are then always able to guide the next person in the picking process to where that scan was for that product.

And immediately overnight, they never lost anything. It was just that simple. And, you know, from that, they're starting to do all sorts of things like, you know, it's integrated with their ERP system. Now they have these heat maps, so they're able to actually space out their personnel more efficiently. 

Natan: So they can do better planning of coverage for big operations and This is a pretty cool story because we're not really playing hard in the WMS space, but you know, my favorite thing is to pick apart three letter acronyms and say, hey, you know, there must be a better, probably more affordable and more dynamic way of doing things.

involves like what tech do you have to implement it with? So that's a cool story. Tulip is a front end for sort of a small scale WMS powered by accurate location. And that gives you the ability to run a warehouse differently and more efficiently. Let's talk about DoubleVerify. I can I can start this story because I kind of pushed it through.

But um, I think I kind of called you at some point and told you this, you know, some of our customers are in luxury goods and when they talk about their scrap, it's not a metaphor, it's actually there's gold on the floor and it could be titanium, all sorts of other precious metals and when you come in and out of those factories, there's actually metal detectors and they have scales everywhere to check what was made versus what was the raw material.

Loss for them is a real thing and worth a lot of money and so the idea was that if you could tag the actual bins. And combine it with, like, pictures that are taken with Tulip Vision, then not only you know where things are, you can detect when they're leaving a certain area of interest and alert, but you can also take pictures along the way.

So, if somebody says, well, this workpiece is here and here, Then now it's verified to be there. So that's what we call double verify. So I thought that was a very cool story and generally speaks to this use case with hyper accuracy. You can do just a whole new level of material flow, which is super important when you're trying to lean out an operation.

Matt: You know, the interesting thing about that particular use case, you know, we were up against basically RFID, you know, that was the competing solution and it actually made so much sense, though, to have quantum RTLS there because, you know, you had all these work orders and they had certain velocity through the production environment, but So, You needed to know where they were, and you had certain work orders that you could tell through your ERP or your WMS or MES, whatever three letter acronym you're using for that particular process.

And you could tell that certain ones stopped progressing, but you didn't know where they were. And so, like you said, these precious metals worth thousands, tens of thousands, maybe more, disappearing effectively from all of your tracking systems. Because when you have RFID or when you have barcodes, you only have a capture at a point in time.

You don't have that real time component. With the quantum RTLS system, you have thousands of these work orders, they're sitting on a shelf. You want to know where one particular work order is. You ping that one, get its location immediately, and now you know, you can walk to it, you can actually go grab it, because it's navigated you to exactly where it is.

And that's just tip of the iceberg of the value that comes from digitizing the material flow and the work order flow in these environments in real time. But also like spatially, not just like step, step, step, step. You have it the whole way through. You have this smooth tracking process of exactly where that work order went, where did it get hung up, where's it going, you know, faster, et cetera.

And it's, it shifts the paradigm quite a bit when you think about those, maybe more traditional manufacturing environments that don't typically have that digital information. Now, now you can optimize your process in so many different ways. And that's going to drive return. 

Natan: Totally. And this leads us right to the next story about driving return, because you kind of touched on Wright's first time story with kind of like the torquing application when you talked about picking application, but I love this one.

This is Wright's first time in a tier one automotive and a really nice story about value creation. 

Matt: Yeah, this is one of the use cases where it's just so clear in the value, so immediate, and it applies to more than just automotive, but this particular customer, they came to us to put the system in place for real time quality control in validation of manufacturing processes.

And the reason they came to us with this problem, they had other systems. I mean, this is a very sophisticated company, right? You know, they've got all sorts of robotics, all sorts of sensors. I think they have their own MES system that they've built, you know, over several decades. So a very sophisticated customer, but they did a, like every large manufacturing company does, you'll have an assembly line with hundreds and hundreds of steps.

And the way you work in an environment like that is you'll do 10 assembly steps and then you'll do a quality step, 10 assembly steps and a quality step, some variation thereof. But generally you're not doing the quality control validation in real time. This particular customer got into a situation where they had made a change to the process.

I think it was a new model or a variation of a model they'd previously built. And so they'd done the change of the process and there was effectively an error in the assembly process and it wasn't caught by one of their quality check. Processes until like the second quality check down the road. That's when the problem became apparent because it's not always apparent right away.

And by the time they had actually caught the error and realized it was an error, they were already into the thousands of vehicles. It was already at the point where there's so many vehicles built and would have been a recall. And it was so expensive to basically undo all the manufacturing steps prior to fix it, that it was actually cheaper for them.

To just build new vehicles and scrap thousands of vehicles. Oh 

Natan: my god, it's like a pre recall. 

Matt: Yeah, that's exactly it. And so they challenged us to come in and just validate their processes in real time. So now, had our system been in at the time, the mistake would have never made it past the operator who's actually doing that step of the process.

The operator would have been informed real time, Oh, you skipped this, or the wiring harness, you've actually looped it around the inside instead of the outside of a carriage or something along those lines. 

Natan: Yeah, that's a great story. We can do a whole episode on right first time and activity detection, how it can change all the things.

And the tech is there, you know, with quantum RTLS, but also computer vision, other types of combined techniques. 

Matt: Well, that's the interesting thing. Right? Like, you know, we've talked a lot about what positioning can do, but it's not just about positioning, right? It's about all of these supporting technologies being deployed together in an ecosystem that's open, where customers can easily do that without having to bring in a system integrator to provide a custom solution.

That's really the enablement for what we do. A lot of these customers these days who may not want to make the tech investment to go all the way to the lights out style factory, but can get all the way there or beyond implementing some of these other technologies. Yeah, 

Natan: totally. We've seen it play out again and again and Kind of back to where we started, that the principles in which we built our platforms and technology products today enable them to talk to each other.

And as a result, lend a good hand to the engineers who actually put it to use and augment their operations. And yeah, we can go on for a while, man, but we're kind of out of time. Thanks so much for coming on to the 

Matt: show. Thanks for having me. 

Voiceover: Thank you for listening to this episode of the Augmented Ops podcast from Tulip Interfaces.

We hope you found this week's episode informative and inspiring. We'll You can find the show on LinkedIn and YouTube or at tulip. co slash podcast. If you enjoyed this episode, please leave us a rating or review on iTunes or wherever you listen to your podcasts until next time.