One brick at a time: How Monumental uses robotics to build walls

0:00Hi, I'm Mike Waitzman, co-host of the Robot Report podcast. If your company wants to reach decision makers in robotics, automation, and AI, you're in the right place. Each week, thousands of industry professionals listen to this podcast for the latest news, expert insights, and conversations with leaders building the future of robotics. This spot could be yours. To learn more about podcast sponsorship opportunities, visit therobotreport.com slash podcast.

0:36Salar al Khafaji: What's up, everybody? Welcome in episode 261 of the Robot Report podcast. I'm Steve Crow, executive editor of the Robot Report. Thank you, folks, as always, for tuning into the program here. I hope everybody's having a great week. Joined by Mike Oitzman, senior editor of the Robot Report. What's going on, brother? Hey, Steve. Well, this episode dropping on 9-11. and this is a significant one this year for us Americans. 25 years since 9-11 happened. Guys, I can't believe it's been 25 years. I mean, my kids have grown up. A lot of things have happened since then, but pretty interesting. One historical fact, at the 25-year mark, quarter-century mark, it's no longer considered a current event.

1:22It becomes history today. Oh, really? Yeah. Interesting designation. I think things like, you know, now there's opportunity for more maybe confidential or top secret stuff to be released at some point with this quarter century mark. So interesting whether we'll learn more about what happened on that day. I certainly remember where I was. Yeah.

1:43Salar al Khafaji: Where were you? I was, you know, brand new acquisition. We had just become a part of a different organization. I was in San Diego meeting the new product managers for the first time in the hotel, getting breakfast with everybody. And when this all happened on the screens, we were all sitting there at the buffet looking up at the screens going, what the heck's happening? And needless to say, they shut down the airport. And there was like 25 of us had to rent cars and drive from San Diego back to the Bay Area because we weren't getting home on an airplane for a week. So it was a pretty wild ride to get home.

2:24That was an easy trip. A lot of people were stuck all over the U.S., right, for weeks.

2:30Salar al Khafaji: Yeah, obviously a horrific day. A lot of horrific things followed that day, right? I was a junior in high school. I'll never forget. It was in Miss Traley's science class when it happened, and we were all sent home. I was a cart boy at the local grocery store back then. that's uh i was supposed to go to work that day and my mom was like there's no effing way you're going to work that day right and we're just just glued to the tv because you're just what the hell is going on um yeah it's you know it's it's obviously i got three younger kids my oldest is 12 she she knows about the event obviously wasn't alive for it she knows about it my two little ones are eight and six they don't even they don't have a clue about it right so it is it is interesting you know the the the future generations what is their tie or connection or how has it impacted their life you know it's impacted all our lives right and i i watch i was telling you before we started recording this a lot of the documentaries that are on netflix and all the other streaming platforms i think it's super important that people watch this stuff so that you know you really don't forget how tragic that day was and then again all the other stuff that transpired after the fact, just horrific, but certainly something I won't forget.

3:43Salar al Khafaji: I'm sure people listening to this won't forget as well, but we have a great show for you today. As always, we'll play an interview later on with Salar Al-Khafaji. He's the CEO and founder of Monumental. We wrote about Monumental earlier this week, Mike, a bricklaying robotic system, but just tell the people what they're going to hear in this interview with Salar. Yeah, it's a really interesting solution. You know, I love outdoor construction robots. And Monumental has built a brick laying robotic system. And we've covered lots of construction robots. We featured many of them in the RBR 50. I think these guys are easy shoe in for an RBR 50 nomination this year for what they're doing.

4:27And essentially they use two robotic arms to build the brick wall. One extrudes the mortar while the other one puts the brick into the mortar and, you know, makes it level. So those are all some of the things that they've solved. Very cool solution. And then they've got two other mobile robots. One brings the bricks over to the other robots and one brings the mortar over to refill the mortar arm. So it's quite an innovative little system here that sort of shuttles things all over the construction site that we're going to learn about.

4:55Salar al Khafaji: Yeah, that's awesome. Check out their YouTube page, right? They must have a YouTube page with some videos on there that we can link to as well. Pretty cool, pretty cool solution. And again, Brianna posted a story about it earlier this week on the Robot Report, which we'll link to as well. We'll also recap the news of the week here in a few minutes. And there's some interesting stories that I think Mike and I are going to debate a little bit about what's going on with these stories. We'll get to that in a second. But do have to plug our upcoming Robo Business show taking place October 20th and 21st out in Santa Clara, California.

5:27Salar al Khafaji: This is the 20th anniversary show, Mike, of Robo Business, shaping up to be a great event. You can get your tickets at robobusiness.com. We're going to have over 2 ,000 people, over 100 exhibitors there, some startups on the floor. And you go check out all the speakers, the agenda. Everything is on the website. Super excited about all the people who are going to be there. I do want to plug this one keynote that we have coming up. It's going to be the ending keynote on the first day of the show where we have put together some great people who are going to discuss how they're using AI in their robotic systems in the real world, deployed at scale.

6:02Salar al Khafaji: How is AI helping them and the customers right now? Where is the hype overblown? And then what does the future hold? So we have Teradyne's chief AI officer, Andrew Davidson, Michael Vogelsang from Collaborative Robotics. We have Amazon Robotics. And we now have ABB, who's going to be joining that AI keynote panel as well. And again, for our listeners, make sure you stay tuned to the end of the show. We did this last week, but we'll reveal a special discount code exclusively for our loyal podcast listeners to save some money on their tickets. So if you want to get involved in RoboBusiness, head to RoboBusiness.com.

6:39Salar al Khafaji: Again, 20th anniversary of the show takes place October 20th and 21st out in Santa Clara. All right, on to the news of the week. And we have you covered with all the latest news and information with our network of websites. Check us out at the Robot Report and Automated Warehouse. Make sure you follow us on all the social media channels. Make sure you sign up for our weekly newsletters as there's no better way to stay informed with what's going on in the robotics world than by doing that. Start with a big story, maybe of the year so far, Mike. I'm not sure. I think this is a pretty big one. Agility Robotics.

7:13Salar al Khafaji: We are finally getting financial information out of Agility Robotics. This is, of course, ahead of the SPAC that is coming down the pipe with them and Churchill Capital. I think they're going, it's either later this year or beginning of next year. I don't think they've announced the date, but the SPAC that we've known about for a couple of months now values agility robotics at$2.5 billion. And the problem, of course, not with just Agility, but with all these other US-based industrial humanoid companies, is we don't know. We have zero information about how many robots they've deployed, how much money they're making.

7:52Salar al Khafaji: And it just kind of creates a lot of uncertainty, right? So ahead of the SPAC, Agility had a file on S4 with the SEC. We now know how much money Agility made in 2025. And it's not great, if we're being honest, right? I don't think it was surprising, but again, we now have concrete evidence or information that Agility Robotics made$1.8 million in revenue in 2025. They had an operating loss of$140 million, operating expenses of$111 million. Their cash burned. They burned$100 million in 2025 alone. They claim they have nine customer sites that have digit robots out there that have logged over 65 ,000 operating hours.

8:43Salar al Khafaji: I think, again, there's a lot to unpack there. But I think the crazy thing is the$2.5 billion valuation. That's roughly 1 ,400 times what Agility Robotics made in 2025. So we learned about this over the weekend, the long weekend. But what was your reaction to seeing this story? Well, again, it's great to be able to peek under the covers, right? And I think that's one of the things we're looking forward to as a few more of these companies go public in this space is being able to follow the numbers. Of course, it doesn't talk about units, but we can sort of back in some of that. It's still not a ton of units yet.

9:24And we see other companies as well, you know, in lining up as to go public. And, you know, we also see other companies that aren't on that path exactly yet, but, you know, beginning to produce quite an inventory of robots coming off the end of the production line. The question is, what are they doing? Right. Where are those robots going beyond just research applications? So I think it's great for us to begin to understand what the business model is going to look like here as things continue to evolve. just interesting milestone. Yeah.

10:02Salar al Khafaji: And agility has for a number of years been viewed as the leading, I think, at least, or, or maybe more real, most realistic level headed. I don't know if these are fair terms, but that's how I have certainly looked at them as I think they're taking the most realistic approach to how to get an industrial humanoid out in the real world. And again, it just doesn't paint a great picture, right? And we hope that it can turn around and we'll discuss this here momentarily. They've explained the two different options for how you can get a digit into your warehouse or your factory. They're talking about all this stuff, but it's such a low number.

10:41Salar al Khafaji: Some people are, again, if you, let's get into the financials, you can either rent a digit humanoid or you can buy it outright, right? So if you want to rent it, robots as a service model, where agility basically retains ownership of the robot, but you're charged a monthly fee. There's deployment fees. There's other software fees. They're saying over the five-year lifespan of the robot, under a robots as a service model, it's going to cost over$500 ,000, upwards of$535 ,000. If you want to buy it outright, again, with that same five-year lifespan, it's just over$400 ,000. So you'll save$100 ,000 if you decide to buy it yourself.

11:23Salar al Khafaji: So somebody took those numbers, and if you just do the$1.8 million in yearly revenue, figure it out. If people bought those outright, that's not even five units. If you're renting them, it's even less. Right. So it's not a great picture. I think Amazon's listed in there as being one of the buyers of some of the robots where they bought multiple units. So, again, it's crazy. The key for them going forward is there's multiple keys. Right. Let's let's see how much money do they close with the SPAC. They're aiming for six hundred twenty million dollars at the close of the SPAC, which if you look at their burn rate over one hundred million dollars, that should last them a few a few years.

12:07Salar al Khafaji: They are hiring people. They have their manufacturing facility up and running now, and they're hiring people there. So I would expect that burn rate, I would expect it to increase, right? But they claim that they have a$300 million order for the new Digit V5 that's coming out later this year. that's on hold because they obviously can't fulfill that number of units now. But they're saying that that$300 million order is coming from a single customer. So I've heard from someone that Agility is going to announce an automotive customer here sometime in the not too distant future. I have that on pretty good authority.

12:46Salar al Khafaji: I don't know if that's the same as this$300 million order. But I would say, right, as Agility continues to file these financial reports, if the SPAC closes and they don't show momentum and that they're increasing their scale and their revenue, that's not going to go well for them, right? So they need to start delivering part of this$300 million order and landing other major customers in a very short amount of time. Yeah, well, and it also shines a light on, you know,$400 ,000 even buying it outright is a long ways from$20 ,000 mark that has been promised for a long time. So again, back to the point that I'm not sure when or if we'll reach that type of price point for this until there's millions and millions of these coming up in production line.

13:39Yeah.

13:39Salar al Khafaji: And I don't think that was them. That's Elon, right? No, that's right. That's claiming the$20 ,000 price point. But I think to your point, and it shows how unrealistic it is. They claim in the S4 that they need to and they can lower the cost of it, right, with the scale, which makes sense. That will happen. That definitely will happen, right? And they're getting into projections, and, again, they're cautioning that this might not happen and things can change. But they're saying they can get up to 800 units in 2027, 7 ,000 by 2030, and up to 25 ,000 units by 2035. from a manufacturing capacity.

14:19Salar al Khafaji: And if they did 25 ,000 under a RAS model, that's$2.5 billion in annual subscription revenue. So again, I don't think that is going to happen based off where they're at now. I think they certainly can increase from where they're at now. I think that's a logical thing. But even the financials are tough to look at. And this has been, you can find all sorts of commentary about this on different social media channels. But what digit has showcased so far is tote moving applications, right? Um, and if you're, if, if one robot costs 400,$500 ,000, you know, break that down, it's about a hundred thousand dollars a year and that's all it can do.

15:01Salar al Khafaji: You know, can, can a person do that better, right? A person that can do the tote moving application in other things at a lower, at a lower cost still. Right. So, again, that's where robotics often comes in. If it can do it better than someone, if it can do it more efficiently, if it can do it lower cost. So I think the math is not really mapping on this stuff at this point in time. But that's not an awesome ROI. No, no, no, no. But, again, this is the first one, right? So again, it's a big moment for humanoid robotics and for the industry to be able to have a benchmark of where an actual company that has deployed a limited number of humanoids in a commercial case sits at.

15:44Salar al Khafaji: And again, they'll have to file the others. If the IPO happens, they'll have to file more of these and we'll be able to track their growth or their stagnation as it happens. So stay tuned on that. And again, I wrote about this yesterday too, or on Wednesday, I should say. you know, Unitree is a publicly traded company over in China. They debuted not even a month ago, Mike, and there was a lot of excitement around the IPO when it first happened. Their stock has tanked 40%, excuse me, 53 % from the high trading point on the IPO day. It's tanked 53%. And China's changing, looking to change regulations for how these hematoid companies become publicly traded companies over in China.

16:30Salar al Khafaji: So things to keep an eye on. But moving forward to another humanoid company, Figure, Mike, you're interested in this story where they've planned a partnership with N-Scale to deploy up to 100 ,000 GPUs on a NVIDIA Vera Rubin platform. And this is to, what, train? I know they have a whole other data project that's going on, but why are you so interested in this story? Well, I think it's, yeah, What I find interesting about this story, and I'm going to write this up as well, Figure reported on their website about this N-Scale partnership. What I think this is interesting, so we know Figure's got a huge war chest, right?

17:10So they do have quite a bit of bank to work with here, unlike some of the other. They're definitely well-financed compared to many of the other physical AI companies. That's sort of broadened the scale here beyond just humanoids to physical AI because, again, the core product in any physical AI company is their model, right? The robot is a robot. We know how to build robots. I don't think that's the difficult part of this. The difficulty is in training the model. And we'll talk about the other aspect of the index solution that Figure has released as well. But I'm sort of working on this thesis that I've been noodling on.

17:50I've been talking to lots of folks about that. I think we're headed for what I'm calling an AI winter, Steve. The fact that I believe that with everybody, not just robotics, every enterprise software company is employing AI now. Across the board, all the software that we touch, AI is a part of it. we're headed for this AI winter where the AI compute resources are going to become constrained at some point in the next, I don't know, two to five years. And then what that means is that everybody's going to be consuming, you know, all of the microseconds of compute time that there won't be any leftover compute time.

18:30And that's going to be a supply demand problem, which is going to drive compute prices through the roof. And I think, well, there's a lot, certainly data center builds are a big topic of the news right now, right? I think there's also going to become constraints on the ability to deliver the actual physical data centers. That's the problem here. Not just that we can't supply all the parts to go into the data center, but that actually a lot of these proposed constructions are going to be shut down and not be allowed to be built because of all of the issues that are taking place. So that's the thesis that there's this pending AI winter.

19:10Salar al Khafaji: So you think that they're getting ahead of this? I think they're getting ahead of it. So where I want to applaud, if this is indeed the strategy here, and I'm hoping to talk to figure to validate this, to get ahead of this, taking some of the cash that they have today that's just sitting in the bank, spend it on building a moat around the castle, right? Which is we'll build our own data centers that we'll own. I mean, this is what all the hyperscalers are doing, right? So Meta's doing this. Amazon's doing this, of course, right? They're going to own the resources. If FIGURE is able to build and own the actual compute resources, when the AI winter gets here, they'll be able to continue to operate, continue to grow, continue to progress, where I think it's going to stop a lot of the smaller companies in their tracks.

19:59And so I think, whether you believe me or not, I'd love to debate this. You got to drop me a line, drop me a DM, because I'm trying to validate whether this thesis is valid or not in the sense that this could be a real problem for across all of the software industry with AI because it depends on hardware now. The compute resources are key, as we all know, to making this work until we get to the next level of quantum computing, which is still a decade away. right we just can't process enough bits so and then they have a couple other projects that are

20:36Salar al Khafaji: related to this right so is it index which is your data capturing uh initiative they're they're claiming they're getting tons of video data every second from the users and they also have helix which is their ai system right so you're saying this is all connected right it's all connected yeah and you know index i think it's a it's a brilliant idea i don't know how it's working yet but it was the cool idea to actually, you know, you download the app on your phone, you put it, film yourself doing activities, and you get paid for it. I was almost going to do it, see if I could earn a few bucks, say I contributed.

21:12I haven't had a chance to do that. I don't know if anybody listening has actually downloaded Index and is using it. You know, drop me an email. Tell me how it's going for you. Would love to hear, but I think it's an innovative idea. So, you know, they've got a couple of innovative ideas here, the index thing as a way to get the data. But, you know, it's also telling that that's where the wall is right now. Right. So they need more data. We hear this every time we talk to any robotics company that's an AI, physical outcome. We need more data. We need more data. That's that mantra right now. Index is one way to solve it, that figures investing in.

21:51And then, of course, once you get all this data, you've got to process it. Right. So, again, the strategy all fits. I just want to applaud Figure and Brett for making good use of the capital that they have. We said, what are they going to do with all of the money that they had taken in? I think this is a great deployment of that capital.

22:12Salar al Khafaji: Let's deploy some robots. That's where I'm at. Yeah. That's where I'm at. Let's just deploy some robots. That's where I'm at. Anything to distract or take the attention off of deploying robots. That's where I'm at with not just them, right? But with the humanoids, let's get these puppies out in the world, right? Come on, let's do this. I want it to happen, right? I love humanoids. Let's get them out there. But yeah, we'd love to talk to figure about this project. Last story, interesting story, fun story here. Dynamic Creatures, Mike, came out of stealth mode earlier this week as well. They're looking to bring characters to life with robotics founded by Mark Thierman, formerly of Boston Dynamics.

22:53Salar al Khafaji: He was the chief strategy officer over there, also joined by somebody over from the Robotics and AI Institute, which also has ties to Boston Dynamics. And we've seen it. So it's not technically a Boston Dynamics spinoff. Boston Dynamics has said that Dynamic Creatures is their official entertainment partner. And we've seen Boston Dynamics be in entertainment applications a lot, right, with Spot and Atlas Dancing. we've seen was it spot on america's got talent there's been a number of other instances as well so it looks like uh dynamic creatures is going to help boston dynamics deploy more applications in in those spaces which include theme parks casinos cruise ships different hospitality settings uh they have their own ai control system for this but uh i wonder what type of characters they're going to come out with here i know there's no word yet there's no pictures no images yet of what this might be, but we, again, what we saw with Boston Dynamics, what we saw on America's Got Talent that, you know, the robots with the dog skin on it, you know, and let's face it, we just talked about Unitree.

24:04Unitree has proven that, you know, the humanoids are good at entertaining, right? All of the festivals that they've, you know, they're making, at least finding places to deploy this, whether there's an ROI, and that's a different question. But the point is that entertainment is certainly one area that the robots can find their way. So why not go all in on the entertainment opportunity here, theme parks and the rest? And remember how popular it was when we had Disney on stage at Robotics Summit a few years ago? It's just fascinating.

24:39Salar al Khafaji: yeah and i'll show you you i was going to make this comment and then you and then you said it i'm not going to share any names but we had disney at the robotics summit what three years ago at this point in time and there were people from boston dynamics in the room watching that keynote there was a whole group of them that were there and one of them walked out of the room and said to me man it would be fun to work with them someday meaning they want to provide their legged platforms to disney in some way shape or form whether it's the quadruped whether it's the biped they don't care right yeah so um you know at this point in time dynamic creatures very early on obviously just coming out of stealth mode they have some financial backers but they are working with quadruped platforms from boston dynamics that they said they're working with atlas i'm not really sure what that means because atlas isn't really out in the world at this point in time I think it said they're also working with Unitary as well.

25:35Salar al Khafaji: Or was it not Unitary? What's the ETH Zurich? The antibiotics? They were working with another non-Boston Dynamics quadruped robot. So they're going to be an independent third party. They'll take the best products on the marketing. I think so. Or some derivative of them. I think this is a great idea. Whether they can execute it, that's a different conversation. But we need to get these robots out in the world more. Get them out in the world more. And by doing that, they'll come across different business models, different applications that help them make money. I remember sharing this video with you guys a couple of years ago.

26:11Salar al Khafaji: It looked like it was at a zoo and there was a group of kids at this zoo. And somebody put like an inflatable, you know, those inflatable Halloween decorations that you can put outside. It was like they had an inflatable costume. It was a dinosaur that they had put on top of a quadruped robot. So I don't know which quadruped robot it was. But it was a T-Rex, let's say, running around and chasing kids at the zoo. And the kids were having a blast, right? So it's, I don't know. I agree. I think this has been an often overlooked market where there is some possible things that can be done. So we'll have to keep an eye on that.

26:49Salar al Khafaji: But wish Dynamic Creatures the best of luck. So, again, make sure you follow us at The Robot Report at Automated Warehouse on social media. Sign up for our weekly newsletters so you never miss a beat. All right, Mike, on to the featured interview this week with Monumental. What are we going to hear from Salar? Yeah, so in this episode, Salar is going to share his insights about how they built the company with this concept of constructing brick walls. And one of the things I learned, and we'll talk about this in the show, is I thought that the brick walls were actually supporting the wall. And you live back east.

27:27So brick walls for me are fascinating, Steve. We don't have brick walls in California because we have earthquakes.

27:34Salar al Khafaji: So I didn't realize that there's still a structural wall and that building the facade with the brick wall is standard practice. Yeah, it's part of the insulating feature. But it's architectural, but it doesn't have to be supportive. So the most important thing for them is getting the wall straight and true. and that's why you pay a master mason, right, is to build the wall straight and true. So, and where the bricks go, you know, all that sort of stuff. So that's what they've figured out. So we're going to learn about their approach to doing that. And the other reason why this is important is because there's a lot of those master masons who are retiring.

28:14So there's a labor shortage here. Construction doesn't stop. And so they've got a real opportunity to deploy automation, enhancing precision. And, you know, they also are looking to expand into structural work in the future with, you know, non-architectural bricks, but the more the supportive bricks. So we'll learn more about that. So let's hear from Salar now. All right. On the show today, we've got Salar Al-Khafaji, CEO and founder of Monumento. Salar, welcome to the podcast. Thanks for having me, guys. Yeah, hey, we're excited to talk to you about your business and what you're doing. So why don't we start there?

28:56So give us an introduction to Monumental. Tell us about your solution for building walls with robots. Yeah, so Monumental is automating on-site construction. We are building robots with the goal of having them do construction jobs. And we're specifically focused on break-claying today. So cool. So you've created a robot solution that automates this entire process of bricklaying and mortar and the entire construction part end-to-end, right? Yeah, we do the whole thing end-to-end. It's so end-to-end to be clear that we actually, for the customers, for our customers, we look like the subcontractors.

29:39So you hire us to build a wall. You don't buy or lease robots. We will come in. We'll deploy multiple types of robots that work together. They do the building. They supply the materials. We build the software in-house that actually runs the robots, the AI models, all of that. And the people that will control the robots are our people. It's all vertically integrated. We're a full stack AI native contractor. So, Laura, why didn't you decide to go after bricklaying first? We know there are other companies in this space around the world, but why didn't you pick that aspect of construction? Yeah, so bricklaying is really interesting because it's a very, very high value task.

30:19And it's actually one of the most labor constrained tasks in most countries. So it's definitely in Western Europe, but also in the US. It is easily in the top 10 job shortage categories anywhere you would look. And so that actually means that like you're really plugging up, you know, an important problem in the industry by solving it. It also means it's actually quite high value, which is important, I think, when you start with robotics, because what you want is basically like robotics early on never works, right? That's the reality of it. Like, you know, day one of starting a company. And so to show that you have like positive unit economics to make your business, it just helps to start with something where people are willing to pay more for it per hour, per break, per whatever.

31:01And as you improve your product, you can kind of like go down market and do like more and more things, obviously. So, you know, there are more reasons, but these are like the core reasons why it made so much sense. And I know that you don't necessarily have a deep background in robotics. Did you have a sense of the state of the technology when you began this path? I know you have the automated movement with the Petra system. You move the mortar with the Panama system. You have bricklaying with a piece of robots under Fleet. How did you come to that arrangement of multiple robots? And what guided you on that choice?

31:41So a couple of things. One of them is, interestingly, looking at the humans. So looking at the humans for us does not mean build a humanoid. It's like see how they behave. And so you'll actually see that on construction sites, a big part of the work is material handling, right? And so the thing you definitely don't want is like have your lead mason to actually be like running around with like wheelbarrows of like bringing, you know, with breaks. Like typically another person would be doing that and you want your mason to be like standing next to the wall, building the wall. That's like the high value thing they should be doing.

32:17That is actually also true for your robots, right? Like it will take time to get them whatever, like down from the scaffold, go to like, you know, get another pallet of bricks, get more mortar. And it makes sense to have other robots do that. And so very quickly we realized you want to have a decentralized architecture with different roles. And you want to also like be able to deploy an arbitrary number of those types of robots, depending on the site, like logistics, the site topology. So instead of building one very fast robot, a lot of people ask me about the speed of the robots. It actually, you know, the thing that makes more sense is like, Like, you tell me how many bricks or square feet you need a day or a week, and I will deploy an appropriate amount of robots to meet that target for you.

32:59Yeah, very interesting. Again, thinking about our audience, which is folks similar to yourself, entrepreneurs, roboticists, machine builders, folks thinking about how to use and deploy robotics. I want to get inside your head a minute in terms of the design of the business. And it seems to me you've built your own custom robots here to deliver against the use case you just described. So talk to us a little bit about that design choice rather than, you know, OEMing some commercial off the shelf industrial or cobot to do this task. You've gone down the route of designing and building your own custom robotic systems.

33:40So I would say that we're very pragmatic and not like dogmatic here, right? Like we were not like, this must be the case. Or I do think to some degree as a startup, you want to control your own destiny. But there's always, you know, we also joke like, you're not going to be like whatever, smelting your own like iron and like, like there's a point where you just like buy off the shelf components. And then, so the question becomes, it's like, what level is the right level for your startup, right? And, but I'll give you one example. Like we actually, when we started the company, we started, I'll explain why, we started building the, like, you know, they're not really arms, they're like mini cranes.

34:14I would argue, and that's actually one of the reasons we custom built them for our use case. But we thought we were just going to buy an AGV platform, like the wheels on the ground. We're like, we're not going to reinvent the wheel here. Literally, we're just going to buy something. This is clearly a commodity component. And then when we started looking, at least when we were looking around, there was nothing that met our specs. And it turned out to not be that hard to actually build in-house. And so we just ended up building it in-house. And so what I would say today is I think for us, there's like peculiar things about construction, which is like most of the payloads are, I would say between like one and 10 kilograms, right?

34:52Like what a human can live. They're not like the microscopic levels that some like small, whatever, like a biolab would use. But your range is quite far, right? Like again, think about a human, you need like a meter of like three feet of like range. So you need to like find the right mix. When we were looking, there was nothing that really made sense for us at a reasonable cost level. And then everything that did had way too much precision. Like they were so expensive because they were like very, very precise systems. But interestingly, we did not care about that part because we're solving it with like AI models and with computer vision and like different loops.

35:24And so again, like we looked at a problem from our perspective and then we're like, it makes more sense to, but we do use, you know, integrate our own like stack. However, we still use our own, you know, off the shelf actuators. Like we're not, again, we're not like, you know, making our own coils or something like that. And when you use the stack that is out there today, I actually think you can very quickly get something working these days that just like without, you know, with a lower lift, let's say like a decade ago, I'd guess. Yeah, I think that's a great observation in terms of the evolution of what vision-guided, you know, positioning has done for the industry and even the whole cobot thing over the last two decades now, how that's evolved from a precision perspective.

36:07Talk briefly about sort of the history here. What generation of the product are you in here? How many times have you sort of iterated on the product design since inception? Yeah, this is a great and also like actually impossible question for me to answer. So the reason for that is because if you look at what our robots, the way they look like, we actually have designed a very modular platform. And that is actually like a very key principle. So we talked about the fact that we have multiple robots. that's true, but they're all built on the same platform. And we're on the third generation of that platform.

36:42So that's the AGV with the wheels. And there's like some sub revisions there. If you look at the bricklaying robot, you know, the core system of it, like the arms, the cranes, we're on the fifth generation of that. But then the end effectors, there are different modules and they're like modular and we can swap them around. The module that extrudes the mortar, we're on the eighth generation. That's been like the hardest. But it's like we've been through so many designs there. And I'm not even talking about like sub revisions. And the teams actually like also operate on different velocities. So like we want to be able to like design and ship and test an end effector within weeks, basically.

37:19And I think you should be able to do that, at least as a small skill. Whereas the AGV platform, we think about that in like, you know, maybe like a once or twice a year type of release. And we standardize there. And like we want to have like safety features in there and like all these things. So I think it's almost like you organize the company around those concepts as well. I think that's interesting when you talk about timing is everything, right? And being able to continually iterate. But this is not like prefab housing or automated construction where you're deploying robots to manufacture wooden walls or drywall off-site, which we've also covered.

37:57But these robots need to be on-site. What happens when Monumental gets to a construction site? Walk us through that process a little bit, because I think we've seen the videos and they're fantastic because you have these multiple systems together, but what does it take to roll that out? Yeah, this is a really great question because we're talking a lot about the robots, but a lot of it is actually not about the robots, right? So let me start with before we get on the construction site. we would get drawings, right? Like we would get some like model or some from the contractor, the architect. Typically these days, that's like a BIM model.

38:37This is like the industry standard in construction. It might be a CAD model. And we load that into our software called Atrium, which is like, which runs this whole process end to end. And we parse it. We might enrich it based on some like things we've learned over the years. And then that already gives you like a sort of prediction. We can like simulate the whole site there before we even go to the site, right? Like how many robots would I need? What are the dependencies there? You need to do things in a certain order, et cetera. Then when we get to site, the first thing we will do is actually map the site, which we can do with normal iPhones or normal cameras.

39:09But we need to get... And before we do that, we actually put some April tags across the site, randomly scattered, but we use them as anchor points, which a lot of robotics people will recognize as a typical... And the unique thing about construction sites, I think really unique compared to almost any other context I'm aware of is, You need to do everything in open space, right? So that is like the really, really hard problem. Like where does that first brick go, right? That's even hard for a human to do. It's very different from, you know, putting a thing in a box where you're looking at the box and you know where it needs to go.

39:42It's like, there's like an open space. I need to put, you know, build a house on the right location. And so a lot of the effort is there. And then what happens invariably is you map the model to reality and it will never fit. And that's part of the magic sauce where we actually like make it work the way a human would work. That's like the important part. And then the robots come in and then they start executing, right? Like only then can we start building with the robots. Is there a division of labor? Are there some tasks that are fully autonomous and others where there are still, people are still required?

40:17Yeah, so we still send operators along with our robots, right? And we're like completely vertically integrated. Like the customer does not need to think about this, but we're sending them along. And over time, this has shifted in different dimensions. So this is actually one of the neat things I'd say about like, you know, selling the outcome is like, in a way, you can plug in every like thing that you've not automated with people. So there's like two things, right? Like what is the type of features you can build and like the tasks you've not automated. So let me give you two examples. When we started the company, when we started doing our first deployments, we could literally not do corners.

40:53Like corners are actually surprisingly hard. So we decided to just like start deploying because we wanted to go to market as fast as possible. We had robots that would do only straight walls. And if there was a corner, a mason, like a human would need to just come in and do it. And that was fine. Like it allowed us to learn and iterate very quickly. Today, we can do corners. We can do very complicated. We can do like deep reveals. We can do all these features that are like expected. But we did them literally like feature by feature. And that was part of our roadmap. The other one is when you're building, which parts have you automated?

41:25So from day one, we were extruding mortar and we were picking and placing the brakes, the obvious parts that people think about. But what do you do with excess mortar? How do you point the wall cleanly? Like that was actually, again, like a human task for a long time. And now we've automated that. We have like a robust grab tools and we have a tool that actually points walls cleanly and ensures that to different degrees depending on what the customer wants. You have to place wall ties. These are still pins that connect the back wall with the front facade. That used to be a manual task. In some cases, it still is manual, but we've also automated that.

42:01So it's like a roadmap, right? And step by step, you're actually automating more on those two dimensions. Interesting. I'm glad you just talked about the concept of a facade because the videos that we've seen so far were exactly that. Facades put in front of an existing structural wall. Are you doing structural walls today? Again, just for reference, I live in California. Anytime I go back to the Midwest in the U.S. and I see brick buildings, I'm like a little kid. I'm like, ooh, look at that brick building. It's so beautiful because we don't have brick buildings in California. You can't build because it's not earthquake structurally sound.

42:42So I'm not familiar with watching brick buildings being built. Everything here is stick built, for example. So I'm just sort of curious. I want to rewind back to this process. You were working with an architect. You know, you hinted at this earlier. Are you doing just facades right now or actually doing structural buildings? What's the differences there? Yeah, so first of all, we don't like work with an architect in the sense that we actually explicitly don't want to like have to like modify the process to like we want to be plug and play. And so like we could just like, you can plug us in almost in any case.

43:18So unfortunately today, this is like me lamenting the state of the world, the way you're sad about seeing no brick buildings in California. Unfortunately, structural brickwork is almost non-existent today. And I think that's a very sad feature of modern buildings for a bunch of reasons, like building codes, insulation, like all these reasons. So a typical building, like if you looked at our videos, most of the buildings, we're building in Europe right now. So in this part of Europe, a typical shell is built out of block work first. So that could be limestone blocks, concrete masonry units, those types of block work.

43:55Then insulation that gets placed against it. An air cavity, again, for insulation purposes. And then the brick wall. The brick wall is a facade in that it's not structural, but it's also not just purely aesthetic. You actually need it to insulate the building from rain. You could screw it up, basically. It's not just like it needs to look pretty. It needs to adhere to a bunch of standards. But the structural part, the way most people think about structural today is like block work that you just do not see. The goal is to do the block work too. Our robots can do the block work. It's lower value, going back to the start of our conversation.

44:34So that's one reason to not start doing it, but we could do it. If you now ask me, what is my personal goal? One of my personal goals with this company is to actually drive down the cost of traditional bricklaying sufficiently so that structural brickwork becomes a possibility again, right? Like there's like a few like very bespoke, like, you know, architectural firms are amazed at that dude, but it's like a niche of an industry. I think this could be a bigger thing again. I think it would make the world more beautiful. I actually think that's one of the things that excite me about deploying robotics into the world, right?

45:05They're not just like about cost savings. They allow us to do things that we have not been able to do for a long time again. hmm in in your roadmap i'm just sort of curious to clarify when you talk about block work which are the the more structural blocks with the interior piece is is there more regulation in in the industry in terms of you know how that has to be placed all of what that you have to fill it with rebar and grout and all of the utilities and all that sort of stuff that that's maybe as you said it's it's more work than the the work that you're doing with with the facade which again Once you get the pattern set, the system can just, boom, it's just laying rows and rows and rows without having to stop and worry about it.

45:51Honestly, I think it's a good question. Is there more regulation? I don't know how to quantify that. I actually would say no. Obviously, the structure needs to be sound. But it's not that hard to get it right, basically. You're literally stacking concrete blocks on top of each other. There needs to be the right adhesive or mortar. But fundamentally, I wouldn't say there's more regulation there compared to other parts of construction. Regulation in the sense that you've got to work with the county. The inspector has to come by and look at it at certain points in the process right before they're insured.

46:28As far as I know, this obviously will depend jurisdiction by jurisdiction, but where we operate and also the parts of the U.S. where I'm familiar with the work, that is true when you're pouring concrete. If you're actually like, you need to tie the rebar manually and you're pouring concrete, that is definitely true. But that is not the part that we're doing. But when you're using concrete missionary units, typically that is not the case, unless you're building a very specific type of structure where that might be true for the lows. But to go back to your question, I think from the robotics side, what I think is interesting is we actually, the way we think about it is we're tackling the harder problem today with the brickwork because the brickwork is exposed.

47:07and people actually care a lot about the aesthetics there to the point that if you, it sounds ridiculous, but if you actually like have a break that is rotated by 0.1 degrees, it is noticeable because like one part of it will stick out and you will like notice like why is that? Like it's just like one break or like multiple ones will be sticking out and will look ugly and people will comment on it and it's actually like, but structurally like the wall is still fine, right? Like it's just like an aesthetic problem. You don't have this issue when you're doing like an interior wall that's going to be like, you know, with drywall around it or like that's going to be like, we'll have like a brick wall in front of it.

47:42So because of that, it's actually like typically a lower paid job. Like you will, the hourly rates are lower, the price per square feet is lower, any way you express it. And that actually also tells you about the relative complexity, at least like the way the market like prices these things. And so like, that's the main reason we've not done it so far. Like we think it's just like there's, more to do for us on the brickwork side before we go into the block work. So last question on the software side, Atrium is your tool that you use to sort of set up and define how the robots are going to lay everything down, right?

48:21That's your sort of supervisor there. In there, does it have, like, I'm imagining these home design tools. You can sort of define the pattern that the bricks are going to have, the spacing, the pitch between them with the mortar, all that sort of stuff. all those things are set up in atrium and you sort of pre-visualize in there what the brick pattern is going to look like including brick color and you know the pattern and the bricks that might yeah that's exactly right so you know atrium is not like a design tool like it's not meant to be designed till the design happens by the architect and they're tooling whatever they want to use but then we load what they've done into atrium and unfortunately modern the models from today that we get from architects, they're never complete.

49:04And so we will talk to them and enrich them exactly with the things you just mentioned, like with the right. And I think where it gets really interesting is again, where model meets reality. So let me just give you one example. You will have two window frames or like a window frame and a door frame in a building, right? And the model will tell you something like the distance between them is going to be like three feet or something like that. And depending on what we're talking about, there might be like five bricks

49:31with some like head joints between them, right? Like some amount of head joints between them. You go to the side and it's going to be like half an inch off. It's going to be like just like a different, like what do you do at that moment, right? Like there's actually different strategies, right? Like you can compress your head joints. You can like start cutting bricks. The bricks might actually have a different size. They're a natural product, right? They're like fired from clay. And so like we have to incorporate all of that dynamically, make the same decisions that, and that's also something that Atrium does, right?

49:57Like it takes, there's the design phase, but then there's also like the live building phase with the computer vision loop that's measuring every individual break, checking if it's up to spec. If it's not up to spec, it will dynamically update the model. That I think is like the very unique part of what we're doing and like unique to like on-site construction. Well, speaking of maintaining that level of precision, whether it's keeping the layers level and the walls plumb and as you said, dealing with that variance between the plan and then the physical reality. Can you talk a little bit about what you're using for perception and localization technologies?

50:34Because we're not talking indoor robots, which can be on a Wi-Fi signal. How are you dealing with sensing and connectivity? So for me, one of the biggest surprises of this company, because again, we went into this like in a very like, you know, we're trying to solve a problem. I'm not, you know, there was very little like dogma, like I want to use this technology, has been how incredibly, incredibly effective vision is. Computer, like, I've just like, we bet on computer vision when we started the company, and I still have been blown away by it almost like every month and every quarter by what we can achieve with it.

51:11I assumed we'd be using lasers and like, I don't know, like all these sensors, like we, and we just, we, and we've tried a bunch of them, by the way. Like, it's not like we didn't want, and like computer vision is just like incredibly good. if you calibrate your sensors, you have to do a lot of hard work to get it right, to be honest, right? Like, especially at the... And so that is like 99 % of our stack, I would argue, is just like vision-based. It's just like we use vision to localize, sometimes augmented by like April tags here and there. Sometimes we don't even need them. We use vision cameras with depth, like stereo cameras, to like see the shape of the brick and understand like where we are in 3D space and do the 3D modeling.

51:49We do 3D reconstruction. to build the 3D model of the site. Again, completely vision-based, occasionally augmented with depth data, but mostly vision-based. And then there's a long tail, I would call it, of sensors that we need for a bunch of reasons. We use ultrasonics to prevent collisions, for safety systems. We need to make sure that our robots can't fall off a platform. So we definitely have more sensors, but for the core product that we're talking about, It's like where the AI loop is running, like where the machine, it's all vision. And again, that's been honestly like a surprise to me compared to like what I expected us to use when we started the company.

52:31Yeah, I have two follow-ups to that. First, you just mentioned putting the robots up on a platform. Are there limits to height in terms of whether it's safety or operational ability? So not as far as I know. So we can operate on different types of solutions. And when you're using mass climbers, mass climbers, so that's the thing that looks like an elevator, typically two bass and then this platform that goes up and down. Mass climbers can be unanchored to a certain height, not crazy high. But then when they're anchored, there's basically no limit. From what I understand, we've not done that, but I think you could go up to 100 plus meters, like 300 feet plus without any issues.

53:17The highest we've built so far with success has been something like 12 meters. So that's close to 40 feet. So with traditional scaffolds, I think we've done something like three stories tall. I would not know a limitation why we couldn't double that today. We do require for traditional scaffolds specific weight ratings and safety ratings. The robots, they're not humans, so you need to make sure the scaffold is built to the right level of safety. Right. Well, I would imagine also that if you're moving bricks around, everything has to be pretty sturdy in terms of weight capacity. The other question I had, you mentioned that you're basing this entirely on vision, but construction sites can be dusty, it can be weather.

54:12how do you handle those types of conditions? Yeah, dust, I mean, I guess the answers are a mix of things, right? Redundancy, you have multiple cameras, you have like, but for example, like dust, we've not seen that much dust to the point that you cannot operate, right? Like that has been, so, but we've definitely, like I would say like sunlight has been like really, really hard, right? Like sun glare, harsh sunlight. I think in some of our first projects, we would be outside. There would be suddenly very harsh sunlight. And then also, for example, aluminum foil insulation that reflects everything.

54:56And then you look at your camera input, and it's just white. It's just only white pixels. And you can have the best computer vision model in the world and it will not work, right? And so it's interesting because you need to learn everything about camera optics, exposure. run HDR mode. There's all these things you need to build that are on a different abstraction layer. And honestly, a lot of it is just engineering. And by being deployed, you just get through all these one by one, right? And you get through the seasons, you get through different materials, you get through different regions. I would say step by step today, our team would probably say painful episode after painful episode, that's actually how you build, I think, a resilient system that scales.

55:38Yeah. I want to come back to the robot family that you've built here. And again, we're on audio. So maybe I would encourage folks to come look at the videos. You've got some great videos up on your YouTube channel. But for folks that haven't seen that yet and are just hearing about this, describe the three different robots that you employ on site and the function of each and sort of what they're doing for the entire operation, maybe how they work together to build the wall. So we have three robots. They're all, as I mentioned, built on the same platform. And so with platform, you should just envision small cars, wagons, with four wheels that drive around, rugged wheels made for construction sites.

56:25The main robot is, we call it Pisa, has two tiny tower cranes, I would call them. I guess they're like three, four feet tall. And so that's six feet tall. And so like they're slightly, I guess, like, you know, they're human sized, they're not humanoid. And they look like, you know, small tower cranes with two arms, right? Like two of them. One does the brick laying, like it picks up bricks and it places the bricks. The other one will put down the mortar through an extruder. And where do they get the bricks? Well, they get it from the other robot called Petra, which is basically a supply robot filled with bricks.

56:58They're ordered in bays. And the bays allow you to actually like pre-program them and have like, you know, different sizes, but also different colors, which allows us to actually like, you know, program patterns and like all these kind of neat features. And they can drive independently, right? They're different robots. So, you know, you can get one that gets you like new brakes while the robot, the other one continues to build. And then there's the third one, which is for the other arm to like refill the mortar. And, you know, the easiest way to describe it is basically a bucket on wheels, right?

57:25It's a bucket of mortar on wheels. It's designed to be able to dock under silos of mortar that can basically dump their mortar in the bucket. And we sized all of these up in a way where you can run them between one and two hours autonomously, right? So the amount of bricks, the amount of mortar that these systems can supply gets you about one to two hours of build time autonomously. I will mention one final thing, which is we have a part that's less visible called Pantheon, which is not on wheels. It's not a robot that we also deploy to each side. It looks like a small box. It's actually the command and control system where all the compute, all the connectivity, all the robots basically talk to that box.

58:03It gives us video streams. We have a version with Starlink. We have a version with 5G, depending on where we're deploying, that actually makes this whole thing work together. And then the other thing I love on your website, and prepping for this, you've got some great vignettes, interviews with some of your operators, and their life now working in bricklaying, coming from other walks of life, so to speak. But again, how are they operating this? they are filling the robot with bricks, right? So they're sort of queuing the robots into the bricks and maybe mixing the mortar. What kind of functions are those on-site operators doing?

58:38Yeah, this is a great question. So first of all, let me just say, one of my favorite things about this company, and I guess surprises, is we are able to like, this is an industry that's honestly like struggled to hire people for a very long time, right? There are all these reasons, but it's just like people do not want to go into construction for many reasons. And one of the things that excites me is we're getting all these people to join us because they think it's very meaningful to build houses. People do want, the impact of the industry is very exciting to them. But this removes the part that they're working with technology.

59:05They're not doing the hard labor parts. They're doing the fun stuff, not the fun stuff that you don't want to do. The role has actually changed so quickly. Every quarter is basically a completely different job. As I mentioned, a year ago, they were basically helping build corners. They were pointing every wall. They were doing all these things that were not automated. Today, a lot of things have been automated, but they're still helping fill the breaks. We typically do not mix the mortar. There's like these big silos of mortar that like just like extruded into like, but they definitely still like help with edge cases.

59:38There's always enough edge cases. They will help debug robots break. Unfortunately, we're still not at a point where they will work for years with no interventions. And so they will either like fix things themselves. They do like light maintenance, but they also liaise with our engineers when we need to do debugging sessions. And I'll give one example, which was still not automated. is like lintels, right? Like you get to like the top of like, you know, like flush with your windows, you need to put a lintel up before you can continue. That's like a human task, right? And the atrium will tell you like, it's time to put the lintel down.

1:00:08And that's like, you know, a human task that needs to happen. Have there been any other technical surprises along the way as you've pulled this whole thing together? You know, again, the interaction between the robots, the staging of the equipment, to setting it up to... I think you gave us a few examples already of the evolution of the system. There's like a million things I can talk about, but let me give you this one. I think mortar is so hard. One of the things I discovered is you can actually, you could do a PhD in concrete engineering only. Like that's how complicated, like that's like concrete engineering is like a subdivision of chemical engineering on its own.

1:00:49And that's because it's like this very complicated, Like, I have so much respect for, like, it's just, like, it's an insane material. It's so magical, right? It could be, like, rock solid. It could be liquid. Humidity affects it. Like, the grains affect it. There's so many factors there. And it's really, really hard to deal with it, to be honest. But, like, we've, you know, I'm very proud of what we've achieved there. But it's not been an easy right to, like, deal with mortar at all. Are you dealing with different types of mortar or different sizes of brick? Yeah, that is actually, like, the complexity that we've took on.

1:01:18So zooming out a bit, the bet we're making, and this goes back to our business model, is we want to be, we don't want our customers to have to think too much about the technology. We want to be like, you're labor constrained. You have this problem. Let me come in with technology to help you solve that problem. And that requires us to basically be plug and play and not be like, no, no, you need to change your mortar now. You need to use these bricks. and like, because imagine if you do that, right? Like suddenly you need to like get approval from the whole chain, like from the architect, from the developer, maybe from like local regulators.

1:01:48Like you need to prove that the new materials are like whatever, like fire safe and like structurally sound. We don't do that. We're just like, we will use what you plan to use as if we're like a human Mason that would just like not complain and deal with it, which creates all these problems for us. Like, so it shifts like an adoption problem. It makes it a hundred percent an engineering problem, which I've always said, this is the part I'm comfortable with. At least I can think about engineering. So we made that decision, but that actually means that you have to deal with different sizes. You have to deal with different...

1:02:17And let me tell you a very quick anecdote again. We spent, I think, nine months working towards our first pilot. We had the drawings. We scoped it. We were ready to do it. But we only learned about the bricks we were going to use, I think, a month in advance. And it turned out the bricks they picked were pitch black bricks. And we were like, wow, that is like... Actually, we didn't even consider black bricks. Let's order a pallet to try them. And nothing worked because our entire computer vision system was literally like just seeing black everywhere. It could not detect edges. It could not position them.

1:02:48It requires like this crazy sprint to get it working. Again, this is why you want to be deployed, right? Like you learn these things very quickly. But yeah, like we have to deal with that variation of like mortar types, brick types. We do all of it. Since you brought it up, let's come back to your business model here. And the fact that you don't sell or rent the robot, right? You sell, quote, walls, right? You're delivering walls as a subcontractor, essentially. So what was your thoughts? Why sell walls instead of robots for this market opportunity? So I think twofold, right? But it starts with the market.

1:03:24So if you talk to people in the industry, and even if you talk to investors, people at least a couple of years ago would all tell me, you're crazy to go into this industry. It's the most conservative industry. People do not want to deal with technology. You're going to fail, right? Basically, the TLDR was like, this is going to not end well. And I actually think, to some degree, they were right in the sense that most people in the industry are not excited to spend hundreds of thousands on this new technology that is unproven. And I think actually for good reasons. So I don't even think, oh, they're idiots.

1:03:58I actually think they have very good reasons to be like that. I would probably do the same thing. If you tell them, you don't need to buy technology, I'm going to be a subcontractor, and I will just sell you whatever you need to solve this, that is literally their job to source subcontractors, to judge their quality, to decide how to partner, to build relationships there, to know if they're going to deliver on. And it shows that we trust our technology, because if I'm telling you, I'm going to deliver you a wall with my robots, you're going to be like, wow, that person apparently thinks they have the technology to do it.

1:04:30Otherwise, I'm going to have a problem, or I will need to pay a mason to do the job for you. I'm going to be net negative. So that worked really well. For us, it also meant that instead of shipping a perfect product from day one, which might have required like a five-year R &D cycle, which I think, again, is a killer for a lot of robotics companies, it allowed us to go to the market much faster, plugging in the gaps with our robot operators wherever needed and still delivering what the customer wanted, right? That's the important part. On that thread, tell us a little bit then maybe some of the pivots you made along the way in terms of trying to figure out the robots as a service model in terms of the original path you took, maybe how you were quoting a job, maybe how that's evolved.

1:05:14Again, I assumed you're not paying for materials. The materials are provided and you're just delivering the execution labor to put the wall up, right? But again, how does that sort of fit into your sales model of talking to the contract, the general contractor or the architect against what your costs are going to be so that they know what they're getting at the end of the day? Yeah, so I think the interesting thing is like, I think, you know, again, like if you think about like startup founders, a lot of them are like spending a lot of time thinking about pricing. Pricing is really hard. It's very strategic.

1:05:48It's very complicated. For us, we never had to really do it because there's just like an industry price to like bricklaying, right? Bricks are actually priced for like brick or per thousand bricks or like per square feet. There's like very clear like industry standards. If you're going to be more expensive, no one wants to work with you. If you're going to be competitive, people, you don't even need to be cheaper. people will work with you because it's, again, like a labor-extrained market. So that was clear to us from day one. I think what's been interesting to us in the last year is what we're discovering is we actually create so much value in the process that we have not been capturing.

1:06:22And we're learning to partner with our customer that we are building the deepest relationships with. We're learning how to basically capture that value or make sure we're like... And let me give you two examples. examples, we're seeing inspections after you build. A lot of companies spend a lot of time inspecting what has been built with humans, but also now with deploying these robot dogs as an experiment, for example. A lot of them are trying this stuff, using these tools, using cameras. You don't need to do that with us, right? We literally have a video stream of every Birk Thau's place. We can generate all these reports.

1:07:00I can tell you the humidity of the mortar as we pushing it down for each like join like there's a uh an amount of data that is just like insane to compare and that all again comes out of atrium right but like we can package it in different ways we can present in different ways so we can even tell you like this is where we think we you know we screwed up on those three bricks maybe the bricks were not good initially or like something went wrong there um we could auto flag it basically the other one is a lot of money is spent uh and this has also been a surprise to me on basically just like let's call it everything around like like surveying sites, putting up profiles, alignment, carpentry around making things to align other things.

1:07:39We could strip away a lot of them. We just want a project where we save the customer close to a quarter million, 250K, just on that category of cost, which is incredible. This is stuff that doesn't even go into the building. It does not help you. It's not going to make you happy. It's not like, oh, I got this marble, whatever. No, it's just stuff that is necessary to make construction high. It's just basically overhead costs that we're able to reduce because we're shifting everything to sensors and vision systems. Well, and this sort of plays into what you mentioned earlier about you started with bricklay, but you're not confining yourself to that.

1:08:18What are some of the next tasks in construction that you're looking to automate? Yeah, so everything I just told you, right? The examples are actually all about localizing things in space. So one way to think about this is the product, the technology we're building, allows us to, any object that's identified a thing, grab it, and place it in the right place in this open space, we think we built a technology stack to do it. So the obvious thing is everything that looks brick-shaped is going to be easy for us to handle. And that's every type of block work, limestone blocks, concrete blocks. And actually, that's a massive category.

1:08:57I think most people know it's like we're talking about tens of billions a year in almost every country. So that is the very obvious one. The way we're thinking about it is we want to own the entire shell. I think that is like you think about a way a building gets built. Well, first you need the heavy machinery, like excavators, ground movers, whatever you need for that side. There's some cool startups that I'm sure you guys are tracking them that's trying to automate that. We're not doing that. Once you have the ground is prepped, something needs to be built. Who's going to do that? That's going to be us.

1:09:32So that's going to be interior walls, exterior walls, insulation panels, at some point probably roofing elements, window frames. We want to build the entire shell until the building is there. And then probably other people will come in to start doing the finishing and putting in your faucets. That's at least not on our roadmap today. But that middle part is the part that gets us really exciting and that we think is going to be very valuable to build and own. And so to get there, what are your plans for scaling? How big is the company now and where are you looking to grow? Yeah, so scaling for us is like there's literally scale in terms of number of robots that you're deploying.

1:10:13That is the most legible access to think about. We've built already over 100 robots. On an average day, we're deploying half of them, which is, I think, so exciting. On a random day, we're on multiple sides. We're now in the Netherlands and the UK. We're in two countries. We're scoping our first project as we speak right now in the Texas area. So hopefully we're going to have some robots there in the coming months. And so that is just more robots and you want them in more places, right? We don't want to be completely dispersed, but in three, four regions for sure. And then the other two axes are amount of autonomy, right?

1:10:48So basically making the robots more autonomous is an important input for us to make this company successful. And then the final one is basically the things we can do, right? So going from not just brickwork to more things. Those are the three dimensions that we're figuring out how to scale without getting everything to break down. Move as fast as you can while still doing a great job. Yeah, one thing we haven't touched on, one dimension we haven't touched on that much is the international expansion. Are there any big differences between the markets in the Netherlands, the UK, and the US? Yeah, so far, I think interestingly, in terms of like the construction side, it's remarkably similar.

1:11:29There's definitely, you know, there's definitely differences, right? So, you know, between the UK has like very, I think, you know, since that big fire that happened a couple of years ago in London, the very stringent fire regulation. And that actually affects, brickwork is not really like affected by it, but like there's all these foils that need to be like put in the right place. You need to think about them, right? They're just like part of the construction process. They do touch the facade. And so like we have to think about them in the process. But fundamentally, bricklaying, you know, bricklaying has not really changed across, honestly, like countries, but also across time.

1:12:00Like, I think you can transfer, you know, have someone coming from Roman times, put them on an construction site today, and they probably, like, know what to do. And that is, like, that is telling about the job. So that part is the same. But because we're, because of our business model, because we are doing the job and we're deploying people there, there's obviously, like, very local, a very different, like, you know, So the type of safety standards, whatever certification your operators need to have, that is very different market by market. And we're navigating that and figuring that out as we speak.

1:12:33There's all these acronyms and cards and exams that you need to make sure to do. And we obviously want to do a great job there. So we're figuring that out per region where we operate. Yeah, you know, I started this conversation off and I live in California. I don't get to see brick homes. And yet I live in the part of California that's in the wildland, wildfire interface, right? So here we do worry about fire every day of the summer. And it's beyond me why we're not using materials other than wood to build our homes today. So I think there's a large opportunity for the future for new architecture designs and construction designs that create buildings that, you know, built out of materials, like you said, like a foundational brick in the future that will create homes.

1:13:15that can survive in these regions in the future. So plenty of opportunity, I think, for you to take this decades into the futures. More of the world builds on those types of materials.

1:13:28Salar al Khafaji: So it's been a fascinating look at your business. I want to thank you for joining us today. We look forward to following your success. And I would remind folks to go check out what you guys are doing on the web. You got some great videos that really show how you're deploying all this today. So thank you. Thank you both for having me. This was fun. Awesome stuff there. I like how we chose bricklaying as the initial application. It's obviously labor constrained, Mike, right? There's a lot of construction tasks are in pretty high value, but it sounds like they're on the way to proving that this can complete a complex task.

1:14:06Yeah, you know, and I was really intrigued with their business model of selling walls as a service rather than selling the robots. So, you know, again, one more great example of an as-a-service business model where the outcome is actually the length of the wall and the number of bricks in it kind of thing. That's how they're sizing the cost for this service. And just another reminder that I just want to pitch that, you know, I'm having a panel at RoboBusiness next month on how to build a successful robots as a service business model. And we've got Rick Falk from Locus on that panel. So if you're attending and you're interested in this space, be sure to attend that session.

1:14:47The other thing is that the other non-technical aspect of the whole solution, which I love, and he talked about this in the conversation, was that in creating jobs, we talked about the fact that the technical resources like Master Masons are becoming more scarce, but still creating jobs for non-technical folks. And I love how if you go to their website or you go to their YouTube channel, they've got a half dozen interviews with some of their operators, how they're providing jobs for folks to work with the robots. They call them robot wranglers. These are the on-site operators, the folks that actually supervise the robot.

1:15:31And there's one interview with one gal who was an elementary school teacher, left that left that profession but now loves her job operating the robot you know i think that's a great example of where robots can actually create jobs um you know for folks that never thought that that was in their future right so so i think that's another thing to celebrate about what they're doing here and still providing economic value as i said you know at the top of the of the interview which is that uh you know they're scoping out the delivery of the actual robot i I mean, wall build, you know, and that's what the architect or the general contractor pays for them to come in and actually build that entire wall.

1:16:14And so, you know, it was a great part of that conversation to learn how they're figuring out how to actually deliver this as a service.

1:16:21Salar al Khafaji: I'm going to have to reach out to Solara. My dream house is a brick house. Maybe he can make a test bed. He can come build me a brick house. Be the first one in the U.S. We can document it. There you go. That's awesome. But thank you again to Salar from Monumental for joining us on the show. Before we sign off, one final reminder about RoboBusiness coming up October 20th and 21st out in Santa Clara, California, talking all things commercial robotics development. We have an incredible group of speakers who will be on our stages over the course of the two days. We'll have over 2 ,000 people, over 100 exhibitors.

1:16:58Salar al Khafaji: And this is the moment everyone's really been listening for, Mike, is our discount code. If you want to save 15 % off of registration, make sure you use the code ROBO15. That's R-O-B-O 15. You enter this at checkout. Again, reach out if you have any questions, but it's going to be an amazing two days. The 20th anniversary show of Robo Business out in Santa Clara on October 20th to the 21st. Again, we're going to be out there a couple of days before, a couple of days after. So if anybody wants to show us around their facilities, show us what they're working on, reach out soon because our schedules are going to be pretty booked here in a pretty short amount of time.

1:17:37Salar al Khafaji: But again, thank you to everybody for listening. Enter the code ROBO15 at checkout. You'll save 15 % off your Robo business passes. Again, hope to see everybody there. And that's going to do it. Episode 261 of the show. Thanks again to everybody for tuning into the show. Thanks to Salar Al-Khafaji, CEO and founder of Monumental for joining us on the show today. New episodes of the Robot Report podcast drop each week. You can find us anywhere you listen to your podcast. Please subscribe, leave us a rating, give us a review. This episode was produced and edited by the multi-talented Mike Oitzman.

1:18:12Salar al Khafaji: Our executive producer is Amanda Buhner. I'm Steve Crow. Thanks so much, folks, for tuning in. We'll talk to you again next week.