The ultimate onsite BIM viewers...no viewer. By leveraging cutting edge, proprietary Augmented Reality technology, Argyle is able to provide BIM right on the jobsite, lined up with the work, allowing much more efficient work and easier communication. Very cool technology and a hint of things to come...
Check out more at https://argyle.build
Hugh Seaton: [00:00:00] Welcome to constructed futures. I'm Hugh Seaton. Today I'm here with merit Thatcher co-founder and CEO of Argyle. Merit welcome to the podcast.
Maret Thatcher: [00:00:10] Thanks for having me Hugh.
Hugh Seaton: [00:00:13] I love what you guys do. So I want to start with what you guys do. Tell us what Argyle is and what you guys do.
Maret Thatcher: [00:00:19] Argyle in its current form brings the building information model to the job site in one-to-one augmented reality.
Hugh Seaton: [00:00:26] So you're an AR company, bringing BIM to the job site. I think you guys have an awesome story and I'd love to start with that too. Tell me how you got to where you are, because there's some amazing wrinkles that , you know, your journey is just awesome.
Maret Thatcher: [00:00:40] Yeah. So I paid for law school on the job site, I worked as a flagger. I did some grout. One of those tiny little Bobcat forklifts for awhile. That was the best one. I did that and I went to law school thinking like, that's the thing I got told that I was that argumentative girl and that this was the way to make my mark on the world.
So did that for several years.
Hugh Seaton: [00:01:00] Obviously you were an entrepreneur early.
Maret Thatcher: [00:01:02] Yes, that's true , that was in my family for sure. It was my mom, my grandpa, all of us had our construction companies. So I had a career as a lawyer and my husband was an architect and he's still my husband, but he's no longer an architect.
He's now a software architect at Argyle. And so our story. This is definitely a labor of love. We pivoted our family and decided to become a startup instead. The kid has been really supportive and, it started honestly, when I was visiting the architecture firm and I tried on a Hololens and I saw the building information model through the walls and they're like, oh, it's just this little pro, it's just this little thing.
And I said, no, this is, this is the future. Like, this is amazing. This is what you're doing. And I think it spiraled from there. I don't know if we knew that we were building this augmented reality company, but I mean, early as 2015, 2016, Logan was teaching himself programming so that he would have a better time moving around his Revit models.
And I saw that, I saw this passion and drive and I. You know, I started encouraging him, you know, why don't you go do some software dev, why don't you go do this full time? And he's like, oh yeah, yeah, sure. I don't, I don't know how we get started in that. Meanwhile, I'm going to my lawyer parties, I'm hustling for him.
I'm thinking that this is the thing like, and you can only be a lawyer who's hustling for this business that doesn't exist for so long before you have to just leave lawyer altogether. So that's what I did. And it happened when his firm had a big round of layoffs. So there goes the XR guy and, now he's my XR guy.
So Logan and I started our company Bevel. We had a partner for a number of years, Simon, we worked on a bunch of VR showrooms and things like that. And then, this construction company in town, in Portland called Anderson construction, reached out and said, Hey, why is AR such a big old pain in the butt?
You know, why is this? We have BIM models. We have AR. Ergo... why can't this just be easy? Why can't we just be running through this? Why can't we be doing quality assurance on our embeds? Why can't we be managing the project in this way? And you know, this is a thing that we've been thinking about forever at our company.
So we said, Hey, there's some stuff we'd like to try. We did that, it was successful. We ran that prototype with a few groups, and then we, shortly after that founded Argyle and turned our attention full-time to, bringing BIM to the job site in this really robust, interactive way that preserves the metadata.
That's really easy. That's actually kind of what it's supposed to be. You know, let's Star Trek version that we were promised.
Hugh Seaton: [00:03:58] Holodeck on the job site. Yeah. what, what was that first thing that you did? So you said you was, a pilot or a MVP or a demo or whatever it was, but something worked.
Maret Thatcher: [00:04:09] yeah, I think it was that we saw the big picture , that they saw the workflow.
I think it was like, we, we took it out to a job site to do a quality assurance of structural embeds. Company is a big concrete self-performed company. So, oftentimes a major project risk is these embeds. And if they're not in the right spot and you pour the concrete, it's a big old problem.
They're tricky though. They're they're small when you compare it to the size of a city block and the current method of doing the quality assurance is you place them and then you use the same plans and assumptions that you use to place them to do the quality assurance on them. And this leads to so much rework just things like this.
Whereas if you could just see this stuff spatially, if you could just see this aligned to the job site in a way that you didn't have to make all of these calculations that you know are rife with human error, you could just do the job. You could just walk up to the embed and say, oh, There it is. That's what the model says that one's in the right spot. Oh, the model says there should be an embed here. There, there is not. So with that initial test using sort of a limited model, a limited scope, we were able to show the workflow in a way that they were excited about.
And then the next job was making it actually work so that you can actually open up your Revit, click one button, have it sent to the cloud, open up your hololens and then have it actually aligned to the job site. So that's the current workflow. That was the big, heavy lift that happened after that initial prototype of just tiny little model, one little area, and manual alignment.
Hugh Seaton: [00:05:54] That is awesome. So I wonder if, you know, an analogy might be the way surveying was done before, where people use eyelines and, you know, they dropped plumb lines and so on to kind of line up. Versus now where you and this may be a bit of a stretch of an analogy, but now you have a laser that shoots straight to where you want to go and you know exactly where things are and you know exactly like where the line is.
Is there, is that an analogy that, that resonates at all?
Maret Thatcher: [00:06:22] I mean, somewhat, , I do see that. Yeah. We're, we're going from an analog to a more digital, but the laser isn't preserved after you turn it off. I think what, what I liken it to and what I'm trying to coin is this idea of spatial project management.
And because you're building information model is inherently spatial. And so is your construction site, much of the management and the documentation that needs to happen could happen on your feet. And so I think of the building information model as the operating system of the job site, maybe that's my analogy.
Hugh Seaton: [00:07:07] I think it's an apt one because that's what we mean by digital twins, right. Is, is BIM turning into something that is functioning and living and breathing with different data coming and coming to and from it.
So as it's used, how talk to me a little bit about how it registers, like what's the process of taking this ephemeral digital thing and lining it up with, you know, steel and concrete, walking away and you come back the next day. How do you guys...
Maret Thatcher: [00:07:35] That is the secret sauce, but what it looks like from the user's perspective, you have to do an initial alignment. You know, you have to tell it where it is. So that initial alignment process is a dollhouse or rather a hologram model of your building appears in front of you. And your first job is to orient that to the job site. And I mean, like north, south, east west, the little dollhouse will kind of follow you.
And then the next job is to place yourself within that doll house. So you grab your little avatar, you put yourself inside, and this is what we call initially... like you're really roughly aligned. It's probably going to look awful, but there you are. A hologram surrounds you.
The next step is to match the holograms to the real world. And this can be done in a couple of ways. If you, like a lot of times when I'm doing a demo, I just do it visually where I'll say, oh, here's a slab edge, or, you know, whatever. Here's a column. The way I would have it done on an actual project is we would tie it to survey coordinates, and that's, I've been hearing the way that people prefer.
Okay. So after that happens, you match the hologram to the real world. You go set one or two anchors and you're set and you go walk around.
Hugh Seaton: [00:08:55] And those anchors, how, if let's say you walk a hundred feet in one direction, do you have to you, is there a way to continue to adjust it?
Because presumably you know, it is with an angle it's it seems like you're right until you get a hundred feet down and it's like, well, we're six inches.
Maret Thatcher: [00:09:10] Yeah, it does. It does pretty good with that, and as you walk there is Yeah, like you say, drift. And so what Argyle does, is it allows the user to either manually correct for that drift, or in future iterations of our product, it starts doing this on its own. But yeah, you can walk, you can set more anchors and our software is maximizing the precision of you, of the local elements, so that close anchors are going to inform my software, as far as like the alignment of the holograms that surround that person, the anchors that are near that person will inform it far more than the anchors that are far away.
And because of that, it can, it's kind of like a feathering effect. It, it works for The messy real world, like a model is perfect. The construction of that model is like a cake, right? It's it's going to look really good. You're going to nail it, but it's, it's not going to be a hundred percent perfect.
It's going to be half an inch here and half an inch there. And that's what honestly, augmented reality had to be able to respond to that type of messiness.
Hugh Seaton: [00:10:20] Yeah. And tell me how you respond to that messiness.
Maret Thatcher: [00:10:23] You do it right now through assigning more anchors and I guess I need more of an example, like what you mean.
Hugh Seaton: [00:10:32] Well, I think, you're saying that that just like you would, if you were trying to kind of pin something down to something real in the world, you know, you don't just put an anchor at one anchor at one end and one at the other and hope it goes well, you tack things down along the way to kind of deal with some warp and weft of what's under it.
Maret Thatcher: [00:10:54] Yes, I somewhat agree with that, but in the ideal use of my software, I don't want users putting out anchors all day long. I think that's not their job. They just have to be using the tool to reference what's around them. So, so that's your first time alignment. You set a bunch of anchors.
Then the next day, it's going to be able to find itself using those anchors wherever it is. The self-healing aspect of it is that it will set new anchors as the time goes by. So if there's one anchor that the software is like, oh, Hey, that is completely not there anymore. There's it gets broken.
I don't see that the other anchors are allowed to inform that. And it can essentially like a scab get new anchors.
Hugh Seaton: [00:11:40] This is great. I mean, the HoloLens, which is your main platform, right?
Maret Thatcher: [00:11:43] It's primarily Hololens, we are actively building our iOS version and hiring genius unity devs. So please send those our way.
Hugh Seaton: [00:11:53] What I'm hearing you say is if anyone has used the HoloLens, one of the cool parts of a Hololens is how it creates that mesh. And so what I'm hearing you say is, it's putting one of your anchors is in the middle of that mesh, but because the HoloLens has created a more complete mesh, it's able to infer other things. If, if that particular point is missing, is that more or less? Right? I'm just sure. There's some secret sauce in there and I don't want you to give away the recipe like that.
Maret Thatcher: [00:12:22] I think you have it right. Only the Hololens isn't doing that.
It's our software that's built on top of... so we're using Azure spacial anchors is one of our inputs. Yeah. We like them. We think that their pricing models ultimately going to be an interesting one, we might have to roll our own, but you know, that's something we keep doing over here at Argyle.
Like I said, when we were, when we were talking before the podcast started, we had to dig in really deep. We started our company November, 2019. So we dug deep and just kinda cleared some tech hurdles. Yeah.
Hugh Seaton: [00:12:58] That's and that's kind of the story of today's tech journey. Anyway, is you use cloud platform services until you think you're ready to engineer your own. Cause you've got to get it off the ground with something.
Talk a little bit about the Azure anchors. Cause a lot of listeners won't know what we, you mean by anchors and there's a couple of people doing it. Right. I think Google was doing it. These they were doing, I don't know if they still are, but talk about what the anchor thing means.
Maret Thatcher: [00:13:22] Sure. Yeah. It's it's you have to tie the real world to data somehow. And anchors were a solution and I know, like you said, a few companies have thought of anchor types. Since we're on Azure, we're using Azure spatial anchors and they are, I guess, yeah, that bridge between tech and digital, they were an assignment saying you are here.
It even looks like a little location bubble.
Hugh Seaton: [00:13:52] But how does it work? What does it anchor to?
Maret Thatcher: [00:13:55] Oh my goodness. That I don't know. It's not GPS. I believe they're using, SLAM.
Hugh Seaton: [00:14:02] That's cool. So the point though, is that some of your tech stack is relying on things that a cloud provider does.
And a lot of the, the magic is what you guys have grown. And again, as you mature, sometimes you're going to engineer back and sometimes you're going to say, you know, the cloud thing is fine. We're going to stick with.
Maret Thatcher: [00:14:19] Yes, absolutely. Yeah. I think we'll be using Azure spacial anchors for a really long time to come.
Hugh Seaton: [00:14:26] Well, let's shift back to how this works for companies and the value that you're providing. I mean, one of the things we like to talk about is what's the problem in the industry that you're solving and one of them you've talked about already this idea of rework, but there's something about AR that I want to discuss with you a little bit, because I think there's something special about new enough technology that no one's sitting around waiting for it. Although it sounds like Anderson was certainly thinking about it. But the fact is you're helping people in a way they didn't anticipate.
Maret Thatcher: [00:14:57] Yeah. That is what really lights me up about this whole thing is, myself, I was so skeptical.
I didn't want to try on a hololens. I was just like, Ugh, no, I don't want to do this. This is going to be weird. I've been to Disneyland and I didn't think it was that cool there, and I'm not going to think it's that cool here. It's just that 3d communication is better. It is easier. It is more comprehensible.
And particularly when you're building 3d structures, I feel like right now we're giving people this ability to communicate in 3d and interact in 3d. And explore a model on their feet in a way that just wasn't possible before. The ROI on that, it's you find your use cases that are very particular to prove it, things like quality assurance is very, I don't know, the low hanging fruit. It's very, it's very nice to be able to say, Hey, we checked and everything out here is within a gnat's ass. It's all good. That's something you can't do with the tape measure and plans, if you do, it just takes all day.
So things like that.
And then I, what I'm excited about beyond the quality assurance application is. That you can interact with the building. I imagine future applications in facilities management, where you're able to just go up to your HVAC, you can click it and say, okay, we know who installed this. We know when we know what kind of material this is and Hey, this is the kind of part we have to order.
Boop, click that link. It's in your cart. I don't know. That's that's the dream is to use the BIM in a spatial way on our buildings.
Hugh Seaton: [00:16:39] I love that idea. I love that vision that you're adding interactivity and not just passive information, although again, the fact that you're able to overlay building information models and geometry, and then the data over the real world and allow people to say, all right, you know, you're, right on target or you're half an inch off or whatever it is. In real time, right there is I think magical.
Maret Thatcher: [00:17:05] It feels like, it feels totally magical.
I feel like the traveling wizard, I get looks on the job site. They're like, what you doing there? Uh, Hey Robocop and, and they crack up, they try it and they're like, whoa, this is the coolest. So I do live for that moment.
Hugh Seaton: [00:17:24] Actually I, you know, and I spent some time in... a little more VR than AR, but it's, it is an interesting thing how both VR and well done AR I mean, in the case of HoloLens, the feeling is similar to VR in how immersive it is. And, you know, it's just more, more real than your phone can sometimes be. But it's interesting how frequently people have an, oh my God moment. Like this is like, it's just different than seeing other kinds of software because it is so experiential and visual.
So have you ever come across this, that the idea that, and in fact, you mentioned earlier, but you know, the visual cortex is one of the biggest parts of the brain. We were kind of wired to figure things out visually as opposed to, you know, numerically or, or even conceptually, is that something you think is really, you know, part of what's making it work?
Maret Thatcher: [00:18:13] Yes, I absolutely, I think it's also part of the skepticism that early AR has received, it's that, you know, we don't need this. We're really good at building. We know we can interpret a plan. We don't need this.
I heard early attitudes, even making fun of clients for not understanding, the spatial relationships given to them when they were given like, you know, architectural documents and for myself, I just think that that's silly.
I think it should be super, super accessible. And augmented reality is surely the way to do that.
Hugh Seaton: [00:18:51] Well, yeah, because you're not asking somebody to learn very much. I mean, obviously there's going to be things they need to learn, but it's the most accessible way to see information. Cause that's, that's how we see it now.
Maret Thatcher: [00:19:02] And going back to this embeds, I mean, when they first told me their embed problem, I hadn't heard what an embed was. So then my next job was to go do a quality assurance using my own software. just doing a big gulp, I had no idea what I was getting myself into.
But it was so easy! I could do it. It was so easy. I could see everything if you'd handed me those plans. I would not have known what I was doing. I wouldn't have been able to make a meaningful project contribution, without a huge lift beforehand. And The way we are needing to move in the industry is, is not to gatekeep, but to use tools that are familiar to educate and not to get you're going to throw away your 2d plans, I still think they're super valuable, but the 3d is going to be a bridge to bring in the demographics, the new talent, the new generations into construction. And it's going to also help with the older generation to communicate with the younger generation about construction concepts. So, I've really been so pleased by the, reception when I go to the job site and I will bring, a hololens.
And one of the oldest guys on site was like, That's the thing for me, this reminds me of playing Oculus with my grandkids and I was like, no way, this, he has an Oculus.
Hugh Seaton: [00:20:27] Yeah, that's cool. Um, and actually, it's interesting. I spoke with somebody the other day, who was talking about how one of their clients decided to adopt a totally unrelated, but adopt technology because as they're looking to attract people in an increasingly scarce market, you know, people prefer to work somewhere that looks like it's modern, up-to-date, you know, using the best technology possible to make their their employees safe and effective and efficient and, you know, kind of happy and good tools do that.
Maret Thatcher: [00:20:58] Yeah. My generation expects instant access to information, and if you're not going to have an up-to-date BIM model as a hologram on the job site, I just don't even know what you're doing.
I just, I can't even, I can't even with you.
Hugh Seaton: [00:21:17] So talk to me about what your vision is. So where do you see the world going? I mean, you guys have your own roadmap, but more broadly. How do you see this evolving over time?
Maret Thatcher: [00:21:26] Gosh, it's spatial project management. I think, someone's going to do it. Project management, as we know, is just such an important part of the construction job site and is something that the field has adopted really easily.
And I think the next step, for that is to tie it to the physical location of the people. And to, I mean, there's so many things you can do with that, from safety to, just document tracking to, you know, like what we're saying, visualization communication.
Hugh Seaton: [00:21:58] That's cool. and the information is visually out there not just kind of in an app that you use. Talk to me a little bit about whether you're seeing and it's early days, but are you seeing that, that this new way of, of doing QA and other things that, you know, your product's being used for?
Is it changing how people do things?
Maret Thatcher: [00:22:17] I think so. And I think that we should also put this in context of how recent some QA checklist processes are for the industry. Like when did that become industry standard? It wasn't very long ago. And yeah, it's this ability that they didn't think about before ittakes a part of the checklist and it makes that part visual.
I think it's changing the way we think we can do things and it's giving me a lot of work to do because they, once I give somebody a quality assurance tool like this, they say, okay, but also I want... and keeping up with that and managing expectations of what our little team can do, is both promising and tricky and exciting.
But yeah, we just need to build that a bigger team and give the people what they want.
Hugh Seaton: [00:23:12] Yeah, it's funny what you just said. You know, you hear people talk about how construction is so slow to adopt technology, except as soon as you give somebody something they like, they're not slow to do anything.
They're the opposite. They're running faster than you are. And you're like, hang on a minute. Hang on. I've only, I've only got three people into, you know, a developer and a half. It speaks to the fact that it is a tough industry to sell into, but not because people do or don't like technology. I mean, obviously there's this, there's a story about everything, there's always someone who doesn't and so on. But I think the real issue is that there are people are just tough on technology that doesn't deliver value...
Maret Thatcher: [00:23:47] And they should be, I mean, who has time for that? And that's my special contribution in this business is I am resident grandmother, I don't want to mess.
And if our software is not easy for me than it is not going to be easy for the users on site. So I love to be that quality assurance person within my own company.
Hugh Seaton: [00:24:12] That's an underrated, but critical point is somebody who can like...
Maret Thatcher: [00:24:17] I didn't join this cause I loved AR and I need AR to work.
No, no, I joined this because AR was disappointing me and I needed it to be better. I needed that. I needed to tell it what for.
Hugh Seaton: [00:24:28] But I think that that speaks though, to one of the reasons why AR can be, has so much promise is that people can see it working quickly, as opposed to other kinds of like analytics software- it takes a little while to get rolling. Then, you know, you need to get information in, you need to gather data for a month or two months or longer, whereas AR people can immediately say, okay, I think I could see how this goes.
Maret Thatcher: [00:24:52] Absolutely. The put a headset on somebody, they say, okay, will you save me three hours of measuring and comprehension.
And like, I know when I'm to install now, like,
Hugh Seaton: [00:25:02] That's awesome. And to the point about quality, like you were saying earlier, you can just see it, like you can say, all right, that's lining right up. I get it. Or, you know what, we're off for a bit let's let's adjust or whatever. Yeah. Excellent. So I think this has been a fantastic conversation.
Thank you Maret for joining. Where can people learn more about Argyle, and try it?
Maret Thatcher: [00:25:24] Yeah. Go to argyle.build. We have a link where you can schedule a demo and the ideal use cases, if we can do that on your job site with your BIM. But barring that we can also do a teams call and just show you what we're seeing, which I got to do with you soon here, Hugh.
Hugh Seaton: [00:25:42] Yeah, love it. Love the idea. A hundred percent. I'll make a BIM model of my apartment and you can do it via teams. You laugh. I've actually done that. I've actually just to be able to try stuff out.
Maret Thatcher: [00:25:53] I'm right there with ya nerds unite.
Hugh Seaton: [00:25:56] We all have our geeky moments. Exactly.
Right. Fantastic. Thank you so much for being on the podcast.
Maret Thatcher: [00:26:02] Thank you so much.