Alex Wyatt: When seven years of platform work becomes seven-week product cycles

23:45Like you can probably see that in the wall, but you're not going to do that in a warehouse or what have you. So if you, you're going to have to drill all those holes first, then you come back and install all the anchors, right? Then once the anchors are all installed, you bang the pictures in, right? And you screw them in. If you're doing it in a robot way, once you've drilled that first hole, you don't have to wait for all the holes to be drilled before the anchor installing robot can come back. You can say, right, I've drilled a hole at X coordinate 52, Y coordinate 12. That one is now available for an anchor.

24:15And so let's say you've got a, I don't know, a five-week process to do the drilling and then a four-week process to do the anchoring. Even if you're the same speed, which you're not, you're five weeks for the drilling and then you're four minutes for the anchoring because that robot can follow along. if you hear what I'm saying. Because as long as they're on the same system, right? And one of the things we spend most of the pandemic working on was this collaboration system so the robots could speak to each other and share information. Because what that gives you is the ability for different types of robots to collaborate on doing different types of tasks.

24:46With August Robotic? With August Robotics. Okay. Yeah. The other way that's really beneficial is you can have multiple robots of the same type that collaborate. and that essentially gives you infinite speed. So for our exhibition robot business, you show up in Las Vegas like most, some nights during the summer or the winter or what have you, we don't have one or two robots running because they just need to get it done pronto. They just need to print it all really, really fast because they don't want to pay extra rent to the Las Vegas Convention Authority. They just want to go. So we'll have 10 robots going and they'll all be doing the work.

25:23And then if you put five on and the estimated completion time is, you know, four o 'clock or 4.41. You put another robot on. Okay, updated. It's now 3.54. You put another one on. All right, 2.51. You can see exactly when it's going to finish. And then at the final minute, they'll do the final, all the nine robots or 10 robots, they all finish at exactly the same time and then they all wander over to the truck, right? Because they're talking to each other. They've allocated that work such that they all finish at exactly the same time in the most optimal manner, right? And that's just with robots that are of the same type.

25:53Once you have that, and then you mesh that with robots of different types where there's dependencies between tasks, like the example before with putting your pictures up on the wall, then you can get some really good power in terms of collapsing different parts of a construction process into one by virtue of this collaboration that we have on our platform, which is, you know, we're proud of a lot of things, but super proud of that robot collaboration piece. and it's not, it was pretty hard to build. It took us a lot longer than we thought, but we're really proud that we pushed through and got it done.

26:26Maybe it would be great if you could sort of reaffirm the advantages to August Robotics on the third way of the platform with the example you just gave. The collaboration is part of the platform, right? And so ultimately, if we're, you've got multiple Lionel robots, we call them Lionel robots because they make, well, they were originally going to make lines on the ground. We call the Boris robot, the drilling robot, because it bores holes into the ground. So we have sort of bad jokes for each of the names of our robots. But ultimately, that fleet management system is a key part of it. So just as much as you say, all right, the robot can get from one point to another, the same fleet management system is used for the Lionel system, as it's used for the Boris system, as it's used for the disinfection robot system, or even if you've got multiple of those robots all working together.

27:15It's exactly the same fleet management system. And so we actually see that as the most important part of the platform of which we speak. So when we talk about the third way, not just a third way to get from one place to another and put different tools on the robot, but basically taking one minute to say, right, you've now got full robot collaboration, right? So, you know, if you were to build a bespoke robot to, you know, hammer nails into a floor and you wanted to build some collaborations that you could have five different robots all hammering nails at a different time. You then need to build a whole new collaboration software in order to actualize that.

27:48We get that done in one minute because the collaboration software is already done. It's already proven. It's already been used on 250 million square feet of area in one area. We've already done 50 ,000 holes in another in a couple of months. That's all proven and it's all there. And so you can just plug straight into that platform. And that gives us enormous advantages as we move forward, particularly in use cases that demand speed. on construction. I'm yet to come across a construction project that's not behind schedule. So, you know, speed's pretty important. Even if they're not behind schedule yet, if they can finish early, they're going to get plenty of bonuses.

28:22So everyone loves speed. Exhibition, like, yeah, I mean, they all want to compress as soon as possible, you know, into as smaller time as possible to set up the exhibition because you pay rent for the exhibition center, not just for the days where Mason comes and talks about his podcast, but, you know, also in terms of the setup. So the more they can constrain that, the better. So speed is a huge part of customers' reasons for wanting to embrace robotics in the modern world. So having that fleet system, that collaboration between robots is the way in which we address that speed pain point. And shifting gears a tiny bit, data centers is a pretty clear tailwind.

29:00Yes. It's obvious to say that I think it's what,$500 billion has been dedicated to the US for infrastructure generally, but very clearly companies like Meta, Google, you name it, have openly dedicated tens and tens of billions of dollars. is it just obvious in hindsight that you found yourself in data centers maybe share the journey of what it's been like working with maybe one of them? Yeah, yeah, absolutely. Yeah, I mean, I think I saw an article in Bloomberg about a year ago, maybe a little bit less, six months ago, was saying that in 2025, it was expected in the US that there was more expenditure on constructing data centers than on offices.

29:45And when you sort of think back to, you know, growing up and whatever, you're always seeing the latest office building go up in central Sydney or in Melbourne, and there's a big skyscraper and there's this and there's that. It's a bulwark of the economy is building offices. And yeah, it's gone down a little bit because of COVID, but commercial construction for offices has been a huge part of the economy for as long as any of us can remember. And the the concept that we're building, spending more on building data centers in 2025 than on building offices in the US just boggles the mind. That is actually quite insane.

30:18And even three or four years ago, like the two lines are massively apart. Now they've crossed over for the first time and that trend's only increasing. And so I think people are probably, they're just underestimating what an enormous sector this is, like in the modern economy. And you give credit to the US government. I mean, they saw that really early. That 500 billion you're talking about, like Project Stargate. I mean, that's only one of multiple commitments. A lot of the stuff is outside of that. I think there was a McKinsey report recently that spoke about, they talked about it as a$7 trillion race or something.

30:52Now, you know, that's obviously over multiple years, but you can cut it any way you want. This is a really big industry. And so in terms of how we got there, I think a lot of construction companies in the US have woken up to this just in the sense that every second new order they're getting. Take our money. We need your tools. I just want a thousand drills. Well, by the way, what are you using these for? They're onto that a little bit. And so it's, whilst I think the general society hasn't yet clocked onto the fact that how much of an industry this is, construction companies have, because they've been supplying the boom as it's grown.

31:32And so if you're working with construction companies to address their pain points, it's only natural you're going to be pulled into the biggest area of growth in their business, which is data centers. And so I think it was a pretty natural thing for us once we started moving in the direction of construction and construction is such a perfect fit for what we have as a robotics platform. I mean, a platform, it's good a lot of things, but primarily we talk about accuracy, being really accurate in a really large area. We talk about autonomy, talk about that collaboration. You know, as we spoke about before, all of those things are really important for, you know, you don't want to, if you're inaccurate in construction, your building falls down.

32:07If you're not autonomous and you need people to look after your robots, when you're not addressing their labor shortage concern. And if your robots aren't collaborating, then you can only go at the pace of one robot. It's not really interesting to them. They want speed. And so the things that we're good at are the things that are really in demand in construction. And so once we went into construction, it's pretty clear that you could end up in data centers because that's where all their growth is. So looping back to what's exciting about the last six months, can you share an example of a customer that you recently signed or started working with?

32:39Yeah, I mean, you mentioned before there's a lot of the big companies in the US building a lot of data centers. We've been really fortunate to be working with Google. And I mean, it's a bit of a wild tale. I mean, they're actually the only company we've shown a demo of our robot to. We had originally planned to sort of make an initial run of a certain number of robots, and we were going to put them out in the market, get a little bit of feedback, and then come back with a second version later in the year. They were so enamored by the demo that they saw. They were like, well, we want all of them.

33:14So our feedback from multiple different sources, that plan sort of changed. But they've been a terrific counterparty to work with. They've trusted us with pretty much all their data centers in the U.S., Uh, and we've only really been working. I think we did the first project at the end of July. Um, so it's only been a few months and we've already drilled like, I think 50, 50 ,000 holes or something like that for them. And that has underlined, um, you know, underpinned them trying to go a lot, lot faster in terms of building data centers. And this is, you know, it's common to, to Google and everyone else.

33:49I mean, Google, fantastic company. They've been the best client. but you know all of them are trying to put up these data centers as quickly as they can and so and you know with high quality and with high safety and so anything that they can do to do this faster is obviously really attractive to them and then the other thing about google which you know really respect is they're just all about safety so you often see a bit of a you know a conflict between speed and safety for google like safety is the absolute priority speed is really important to them financially, but they never compromise on safety.

34:24And so the idea that you don't have people, you know, having these really heavy pieces of equipment, like right near their shoes and, you know, they could drop them, they could get shaking, they could, they could drill through their foot. Like there's all sorts of things that you can do and that now people are nowhere near the, near the drills and it's all done by robots, um, is something that, you know, I think really appeals to them from a safety standpoint. So they've been terrific thus far. I've really enjoyed the journey with them. And we do see opportunities to automate other parts of the data center construction process together with them as well.

34:58Right. Congrats. They're a great client. And Anthropic platform, one of the fastest revenue ramps in history. And the reality is that there are billions of computers around the world already distributed with users on the other side of the computer who are ready to use their products. that isn't the case with robotics. And there is this journey of, I need to build the thing. I need to strike the contract. I need to do a pilot. I need to build trust. I need a ramp for that specific use case. I need to ramp up other use cases so that we have collaboration within. And so I think it's fair to assume that we won't see a revenue ramp like Anthropic.

35:38And I think one of the failings of the graveyard era is the promise that there could be a huge revenue ramp, But the reality is that we're in the physical world. And so I'm wondering, what are the enabling technologies that you have control over or the enabling technologies in the market that will help accelerate the revenue ramp, help accelerate the distribution of robotics? Yeah, I mean, I suppose I'd challenge it a little bit in terms of the inability for really fast revenue ramps. I mean your point is valid in the sense that you know a software company like if you're selling ads online you can literally ramp that to everyone on the internet you know the next day and you're not gonna be able to do that with a robot but I think we actually feel really optimistic about the pace of revenue growth because it is harder to get in so it's a lot easier to write a you know piece of software to do something like barriers of entry are a lot lower than to build a robot to be able to actualize physical tasks.

36:38And then on the basis of this platform, and when we're looking at, you know, like our Boris business is reaching a point of inflection where that's sort of going to, you know, they talk about a hockey stick, there's a bit of a hockey stick, but we almost look at it as multiple hockey sticks layered on each other. Because once you've got the platform, you can throw off a new business every three or four months. And indeed, we don't even need to, it's a six month process to do that. It doesn't have to be sequential, one robot, six months, one robot. We could build three in six months. It's just a process where you have to iterate prototypes.

37:05We just build three off the basis of the platform. And so if you've got all these different businesses, all hockey sticking, and you add them all together, your end result is a pretty steep revenue wrap, right? Which is going to, in many ways, be a whole lot steeper than software companies with competitors and what have you. And so I feel as though the proposition that you necessarily have to grow revenue slower in a hardware robotics company than in a software company, I challenge that. I challenge that after you have a platform. before such times you have a platform, but once you have a platform, I think, and that's, you know, ramping back to your very first question about, yeah, where's your head at right now?

37:41Sort of feel like we've spent six or seven years building this really cool piece of technology. And now the fun starts, right? This is about like, how are we going to use that? Not just in terms of ramping revenue, but in terms of doing more and more and more useful things in the, in the society. And it's, it's not going to be linear. It was just a whole lot of investments. It gets where we're at now. And now that inflection point is not just an inflection point for each little business, but layering multiple inflection points all on top of each other. So we feel optimistic about the possibility of ramping revenue really quickly.

38:12But it's not going to be one business that generates$2 million of revenue, but it might be 10 different robot businesses that all generate $300 million revenue. There's your fast revenue ramp. But it's just a bit different, if you get what I'm saying, in terms of the way in which you do that, because there are physical constraints. Yeah. I mean, it's very similar to eucalyptus and generally speaking, sort of the house of brands where there'll be some use cases that just explode and there'll be others which are enabling that one thing to explode. And they're accumulating advantages along the way so that each build layers on top of the next.

38:44Exactly. And becomes easier and easier and easier. Exactly. And I think, yeah, like the eucalyptus story, as you build more things, more of it can become generalized into your platform. And so therefore the next task is going to be shorter and shorter. So what might be a six month process now might be a four month process later. Yeah. The other thing that you do get with the revenue stuff is that, I mean, certainly in our case, we look at our robot collaboration as a bit of a competitive moat in the sense that if these robots are going to be talking to each other, and then that is going to collapse workflows, No one's going to use a robot from outside the ecosystem because they're not going to be able to have access to that particular benefit.

39:22Right. And so that's another way in which you scale revenue really quickly is when you build, you know, Apple's done it. I think, you know, all the social media companies many years ago did it. If you can build these sort of network effects such that it's really very difficult to come off the system or, you know, customers are losing a lot of benefit. That's another way in which you grow revenue quickly because everyone's within that same ecosystem and is incentivized to remain within it. And so for us, that's an important part of the story as well. I'm curious just to learn about, to the extent that you can, some of the decisions from the platform point of view that you look back in hindsight and you go, damn, I'm glad I made that decision.

40:01Oh, we can think of a few opposites as well that we had to redo. But look, the main one that we often talk about in that regard is in terms of, it's getting a bit techie, but in terms of localization. So, you know, we're not the only robot company out there. We're not the only robot company that wants to localize to a high degree of accuracy. All the others use essentially off-the-shelf surveying equipment. And, you know, we obviously considered that really deeply. It's really tempting to use an off-the-shelf surveying equipment because it's already there. We took a decision early on to build our own.

40:33And we did that for a couple of reasons. The main one was they're really expensive. I mean, some of these high-end total stations that you need to run some of the competing robots, like$150 ,000 for one of them. What's tempting about it then? What's tempting about it is that you don't have to develop a whole new product that's already, you know, I mean, the surveying industry has been doing that for years. They've built these things to be able to, you can just put your robot there and it'll tell its position within three millimeters. I mean, it took me years to get to that point without one. Yeah.

41:01Right? But what we decided to do is essentially to build something similar to the surveying equipment, but strip out all of the unnecessary functionality and just do a point-to-point localization, which means that we can do that a lot, lot, lot, lot, lot cheaper. And also we can do it in a way that doesn't rely upon continuous connection between the robot and the total station, which means you're robust to a really busy environment. So, you know, all the competing robots in the industry, if a robot, you know, if Mason stands between the robot and its total station, the robot just sort of loses the plot.

41:32Can't deal with it. Spaz is out. I need to check in 300 times per second, or I don't know where I am, right? Or if the robot goes behind a pillar and the total station's the other side, again, just loses it, right? And these days, construction sites aren't that clean. There are people walking around, there are forklifts, there are pillars, there are heaps of obstacles. And so we took the decision if we wanted to be robust to busy work sites, we wanted a sort of a point-to-point system, and the total station wasn't the optimum solution for that. All these sort of, they call them RTS, robotic total stations, because they're checking in so often.

42:07So both for cost reasons, so that we could include them as part of our system rather than sort of directing someone to like her and say, go buy one of these things for 150 grand, both for cost reasons and then also for this sort of point-to-point robustness for in busy work sites. Yeah. That's the best decision we ever made. Now, you know, there would have been times on the journey I would have said, geez, we should just use the damn total station as it gave the wrong result again and again and again. And you sort of work through that R &D process. But now that we've come out the other end, we are providing as good, if not better accuracy than these total stations.

42:40And we're doing it in a much more effective way and a much lower cost. Obviously COVID was a delay, but perhaps can you share a bit more about the agony of that experience? And like it's taken longer than you'd hoped to begin with, much like every other startup in history. but I'd love to learn a bit more about the battle-hardened version of August Robotics and what you had to go through to get to this point. I mean, you have a lot of founders that come on your podcast. I don't think any of them would launch a product, go really well, it's a nice revenue coming in, and then they have 23 months of zero revenue after that.

43:18That's not on any of the playbooks about how you're supposed to run a startup. And, you know, I think we were, I mean, it was really hard, but we're also really fortunate to have, you know, I think really supportive investors that, you know, were aware of what happened and were willing to support us through. I mean, I think part of it also comes, you know, personally as a second journey, like, you know, this is the second time founder, you have a little bit more perspective. I'm not sure I would have been able to handle that first time around. Like it's a pretty severe curveball. You release your product and then your industry gets banned.

43:51So yeah, it was really hard. The hardest thing about it was we didn't know when it was going to end. And I think a lot of people in the community, you know, we sort of put those COVID years to one side and sort of forget they happened. But if we sort of take our minds back to there, you know, every three or four months, it's like, oh, it'll end now. And so then we'd gear up to start our business again. And we'd, you know, we'd get all the robots out of storage and then Omicron had hit. And then, oh, you know, and then something else had hit. And so it was, if we just know one, there's like, well, you're not going to get any revenue for 23 months.

44:21I think it would have been a whole lot easier than thinking every six months it's going to end next month and then it didn't and so that was really difficult yeah that's agony that is such agony it sucked on on the positive side I mean clearly I wouldn't want it to happen again yes but if you're sort of looking for silver linings on the dark cloud what it did mean is we had this awesome team of engineers and they would have been I think you know distracted in terms of serving these customers and making the robots work and what have you in the exhibitions field. They couldn't do that. So I was like, all right, what are we going to do?

44:53Because at the time when the exhibition started, we had exclusively a service model. People tell you all the time, you know, robot as a service, what have you. One of the main reasons people do robot as a service is it's much easier to get to market faster if your audience is someone with a PhD in robotics to operate it rather than a general member of the public. It's a lot harder to build it. And so essentially when it became clear the pandemic wasn't going to go away after a month, like we all optimistically thought at the beginning, that was the task. that we doubled down on. He's all right, before this pandemic ends, we're going to make this robot, I mean, the mantra in the office was, your mum has to be able to use it, right?

45:25Because prior to that, there was probably four or five people who could run the robot. And you know, the robot would do a great job, clients were all delighted. But if we didn't have one of those four or five people there, you know, panic stations, right? And so what we were able to do is it did give us space to really focus on a fundamental building block, which then unlocked a lease model for our exhibition business. so the vast majority of our business now is lease business. So people just grab the robots, we give them a training course and they run them. We're not there. We're not doing service.

45:53We don't have to drive trucks. We don't have to fly people around. We just send the robots there. We have some wonderful clients at all these exhibition centers around the world. If they have a problem, they call up. But ultimately, we unlock that lease model. And I'm not convinced we would have had that level of focus to crack that really serious problem had we not had the benefit of the pandemic. the other big thing we did in the pandemic was the fleet system, so the robot collaboration. So we came out of it with a legacy. A lot of startups in our position might not have had that luxury. We had some supportive investors that got us through.

46:25But yeah, I hope that doesn't happen again. Fair enough. Me too. For humanity's sake. Well, I'm not talking about COVID. I'm just more talking about losing all the things. I know. But yeah. Clearly, if we're talking about it for humanity's sake, I mean, there are a whole lot worse things that happen in the pandemic than - Oh, sorry. Yeah, of course. I can't be using a little bit of revenue. Yeah. And I obviously know that this is your second time, second term as a founder, if you will. What lessons were you leaning on to keep the team motivated when it is so stop start? Like it's almost as if there's that rule of like, generally speaking, if you're going to let some of your team go, don't do it in like 10 % and then 10 % and then 10 % because it just destroys morale.

47:09And in a similar way, it's like, guys, we're a month away, everyone rally, and then stop, start, stop, start, stop, start. Like, how did you sit with that as the leader? Yeah. I mean, it was obviously it was difficult. We were fortunate in the sense that given we'd only launched the product about three months before, we didn't have a whole heap of sort of commercial infrastructure in place. you know, the classic playbook and, you know, subscribe to that playbook is your first sales should be done by the founder and the key management team and what have you. And so we were still very much in that phase when the pandemic hit.

47:44So we hadn't hired a sort of a big sales team or, you know, a big commercial deployment team or anything like that. It was essentially, we were doing it ourselves and we were learning, you know, we had some engineers that had built the robot that would fly it to then run them. And so it was, it was all still our core team. So it wasn't really that we needed to sort of, cause there's no way you could have sustained a commercial team through the pandemic? Or were they going to be selling? So we would have had just sort of, you know, I think we had one and that was sort of pretty obvious, right? That that had to sort of end when the pandemic became quite permanent.

48:14But the majority of the team was the team that had worked on building that same robot. And so we're able to motivate them. It's like, all right, well, we did build this really cool robot. We did have a success. Like everyone loved it. We know we can build really cool robots. Let's now judge ourselves on our technical progress rather than our commercial progress, because the commercial progress right now, there's not much we can do about it. We didn't create the pandemic, you know, it just sort of happened. And so we were able to say, all right, how are we going in terms of our progress towards this?

48:42You know, your mom has to use it. How are we going in terms of our progress in terms of the robot collaboration? What is this going to open up in the future when all this mayhem ends? And so we're able to sort of focus back in on those sort of micro wins, if not the commercial ones. Now, of course, that was easier for most of the team than it was for me. You have to go to your board meeting and explain, again, that you had no revenue for the last six months. It was, and probably didn't have as many engineers join the board meetings as you would normally do in other circumstances. But yeah, that's something I was able to sort of carry that load and almost ring fence our engineers from that and have them focused on technical milestones rather than commercial ones.

49:22And I think I read, maybe I'm wrong, but I think I read that you're launching office in Australia. Yeah, absolutely. Why are you returning to Australia? I'm really excited about that, first of all. I mean, the plan is we'll be doing that, I think, in January. Okay. When we open that up. We've never had people here. We've got a great client here. Shout out to ICC Sydney. Awesome client that have been running our robots for a number of years. The line of robots running in the exhibition centre in Sydney. but we're building a data analysis, artificial intelligence and data security center. And we were, you know, looking for a variety of locations where we'd put that.

50:04I mean, the pros and minuses of a lot of places, I mean, the US would be pretty obvious, but, you know, we've got a lot of manufacturing and production in China, like most companies do, you know, building sensors and supply chains and what have you. We were really attracted to Australia. You know, one of, you know, we've got an obvious connection here, very attracted by talent. or attracted by the fact that it's in the same time zone as our factory where we build things. Federal government and the R &D subsidy, that doesn't hurt either. And so, yeah, I mean, we're going to be building an office here for a lot of software engineers essentially to work on pulling all the data out of our robots and then using that data in a couple of different ways.

50:44So one is in terms of the obvious one, in terms of data analytics and just making a lot more of it than we do right now. Right now, we just generate all this data and we don't really look at it. I mean, we look at a bit, but we sell 50 ,000 holes. But in terms of there's extra layers of data that we can get a lot more insights out of that we haven't really looked at thus far. So clearly, there's a really big piece there in terms of leveling up our game in terms of data analysis. And then the other big thing with data is you put that into AI models. And so we've got some fantastic things that we can do.

51:13I don't want to talk about them too much, less that they're all ideas at this point. but that using artificial intelligence to increase efficiencies across the whole system, using artificial intelligence to essentially enable the robots to set themselves up. So rather than having to train someone, if the robots are talking to the individuals and telling them what they need, you don't even need a training course. Just the robot will set up and set itself up with some help from the human. And so there's all sorts of different things that we can do with AI. and we decided that, you know, putting that in a deployment office or what have you, sort of divert the focus a little bit.

51:50We wanted to just start maybe sort of eight or ten engineers and they're just sitting there working on these data programs, this problem, working on AI and going really deep into that aspect of things and something I'm really passionate about, artificial intelligence grad from Melbourne Uni back in the day and yeah, really excited to make use of what is a burgeoning scene in Australia for artificial intelligence. And we think it's a perfect location and excited to get started in January. Speaking of back in the day, when did you build your first robot? Well, for all this robotics. No, you personally.

52:25Oh, me personally. Oh, man. Wow. If you could call it a robot. My first robot was I had a friend in Sydney that had an apartment. and I can't remember what it was, Bronte or something. And he had a pigeon problem on his balcony. Okay. And so these pigeons were pooping on the balcony all the time. And he tried spikes and lots of stuff. None of it sort of worked. And the only thing he worked out that he could get rid of these pigeons was that he'd take a water pistol, like one of those super soakers, and he'd shoot at the pigeons. Wow. And that would scare them away. And, you know, it was ethically okay because there was just water and the pigeons weren't damaged in any way, shape, or form.

53:01But then he would go to uni for the day and he'd come back and no one was firing the water pistol. So there'd be bird droppings all over his balcony. And so at that point, we came up with the idea of essentially attaching a water pistol to a bucket of water on his balcony with a motion sensor. And every time it detected some motion outside, it would like spray outside the balcony and front away all the pigeons. And so if you could, I think you can call that a robot. Without a doubt. We essentially had a robot firing water pistols at pigeons on a balcony in, I think it was Bronte or something like that back in the day.

53:35So that would have been my first robot. And I guess to take you back to the beginning of August Robotics, what was the first robot that you guys built there? First robot we built was essentially we bought like a little base on the shelf and we had a big table. There were four of us that essentially built the first robot and we had a big table and our four tables were put together and just this little base that used to go from Marshall's desk to Patrick's desk and from Patrick's to Xiaodin's and we then measure its distance. And essentially it was, you know, that was about, I mean, the whole localization system that we built later will tell you where your robot is.

54:12But if you don't have the fidelity in your wheels to be able to move to that required level of accuracy, then knowing your accuracy is irrelevant. And so ultimately the very first robot we built was essentially some piece of metal on four wheels and trying to tune this locomotion. So it had 360 degree movement at a really fine level of fidelity to get from one person's desk to another. We had little targets on the desks and they'd move around. And, you know, obviously we did on the floor as well. But the first robot was very, very, very basic. Just a couple of wheels and a metal sheet. There's a very clear, like, arc in your life with robotics, but, like, what gave you the courage to go again?

54:50Oh, geez, that's a good question. I think I've always wanted to start a robot company. Yeah. So, I mean, I always saw... Since shooting pistols. Shooting pistols, indeed. shooting at pigeons um i always saw robotics as the as the main use case the best use case the killer app for artificial intelligence in the sense that i mean i'm as excited as anyone about ai and generative ai and what have you but the reality is that all of that is still in the digital world there's no you can't bring ai to the physical world without a robot and equally a robot isn't a robot without artificial without artificial intelligence i mean it's a what's a robot without artificial intelligence.

55:29It's like a remote control car. You know, we've had them for many generations. A plane. I would argue a plane is partially a robot ready, right? In the sense that - You've got, yeah, auto. Autopilot. Yeah. Yeah. I mean, in any categorization, there's always things on the boundary. Yeah. And yeah, I think a plane, absolutely. Yeah. It's halfway in between, you know, a piece of equipment that's operated by a pilot and a robot. Yeah. Because they do go away. That's a robot. But yeah, I've always thought that as an artificial intelligence enthusiast from when it was desperately uncool and my Greek mother was, why don't you do something useful with your life, this silly artificial intelligence thing.

56:10I always thought that robots were that killer app. I always wanted to be part of that. And it's been waiting for the right time. So I had a great experience in renewable energy. In my first startup, absolutely loved what we did. I did actually have two ideas at the time. So one was a renewable energy business with a really good friend of mine who was an expert in renewable energy. The other one was a robot business. But at that time, it was just too early. So I did talk to a few people about it, potential investors. And, you know, their first response was, you know, R2, D2. They weren't really taking it seriously.

56:43There wasn't much data. The sensors were expensive. People just weren't really ready for robots. I mean, we're talking, you know, sort of 15 years ago. It was just too early. The economics didn't stack up. So I was sort of, you know, I mean, we're really happy. The company we did last time was great. It was a great experience. We had good success, worked with a lot of great people, built a really good company. But yeah, I mean, even then I was trying to start a robot company. And so it wasn't the courage to start again. It's like, is the time right yet? Is the time right yet? Oh, I think the time might be right.

57:13Let's do it. So it was more the opposite thinking. I've always wanted to do it. And it was just about waiting for the world to catch up with my aspirations. And how did you make the decision that the timing was right? It's one of the hardest things. Yeah, it is really hard. Or perhaps how did you decide it was, I guess you've already answered why it was wrong. Yeah. Yeah. I mean, I guess the, yeah, I'm just sort of trying to be quite thoughtful about it because it's, I mean, there's such a mishmash of different factors. So part of it's personal, you know, are you sort of free and ready to go and, you know, you're in an ability to be able to start a company with all the emotional baggage that you know that entails, particularly second time around, like, you know what you're in for.

57:54so you've got to be ready at a personal level you need the sort of fundamentals to be there and you know i had an eye on those fundamentals for the past decade and you know it was clear when we started all this robotics that you know sensors had come miles down in price processing was so much cheaper than it was previously it could be so much faster it was faster to the point that you could get the speed of movement that you needed to react in a way for a robot robots that you would have had like early 2000s that could move but they'd sort of be moving in slow motion, right? And it just, it just didn't meet the threshold of what you needed for a commercially useful robot.

58:28So I think there was a fair bit on the, there was a personal side, there was a fair bit on the technical side. And then I did, you know, started to talk to a few investors and previously they would joke, you say robots, and suddenly this is something they take seriously now, right? And so if you've got, you know, the personal ability to go for it, the sort of technology is in a place that it's ready. And then there are people that, could be convinced to fund your adventures, that makes a pretty good time. So those three were ticked off. It was like, okay, time to go. No time like the present, let's do it.

58:59And we started doing a lot of that development with the supply chain in Southern China and then expanded through to Europe and the US. Now it's almost exclusively US for the drilling business and for the exhibition business. We're all over the world, but the strongest business is in Europe. That's exciting. Sorry, did you say you have an office in the US? Yeah. Yeah, absolutely. So Germany? Germany, US. We're in the US? Sorry? We're in the US. We're in Las Vegas. Okay. Las Vegas being in the trade shows. I mean, I've been in the US now for almost three years. And yeah, I mean, we've got Lionel Business there.

59:39We do a lot of trade shows. And it's been really helpful in terms of launching the Boris business, like our collaboration with DeWalt. They're a big, iconic US brand. Being able to be close by to them, collaboration with Google, who have been terrific, makes a huge difference to be there. And so, yeah, US is, you know, absolutely our key market. And then, you know, obviously in the Lionel business, everywhere there's exhibitions. But, you know, thus far Europe's been the most successful, but we're seeing some good growth in the US. And we've got robots elsewhere. We've got them in Korea, got them in Dubai, got them in Sydney.

1:00:16got them in Singapore got them in Hong Kong so they I think we're in Mexico as well so you know we're in a lot of different countries with those robots that's the other advantage of the lease model you don't need to start an office everywhere you go you can send the robots there you train them and then people can use them so the line of business is genuinely global business and at the moment our business for downward drilling is exclusively in the US but excited to bring it not just to data centres but warehouses and other things in other countries pretty soon to finish things off i tend to ask maybe share one or a couple predictions of where august robotics will be next year so when we next have you on yeah next year and the next year maybe five years you can go 10 years you can define the timeline very good we're very brave to be doing a podcast and committing to that in 10 years from now but um no no you can share it yeah absolutely I think the cool thing about Wild Hearts has just been having people come on the podcast over time.

1:01:17Yeah, absolutely. So you have this memory lane of their evolution, both as a founder, but also as a company. Yeah. And this is just one of those questions where you can easily go back on and say, how'd you go? Yeah, no, absolutely. I'll preface that with just sort of a more general reflection. That's cool. And then lean into that in terms of where we think we're going to be. I mean, if I look at robotics more generally, and one of the things that keeps all of us motivated through good times and bad and what have you is that it just feels to me that it's inevitable, right? I mean, last night, I was driving back quite late at night in Sydney, and they're doing some roadworks.

1:01:53People have put out the witch's hats. There's a couple of guys digging something as a semi-trailer drives past two meters from the guy that's doing the work. It's two o 'clock in the morning. Do you really imagine that someone's going to be doing that in 20 years from now, or is that going to be a machine? There is no world where I can consider that that's not going to be a machine. It's not going to be a robot doing that work in 20 years. And you look right throughout the economy at these, you know, we call them the triple D, dirty, dangerous, and dull jobs. There is no way that those tasks aren't going to be done by a robot.

1:02:22They're going to be done better. They're going to be done cheaper. And people want them to do it. That guy doesn't want to be there operating that machine at two in the morning with semi-trailers throwing dust at him. No one wants to be doing that. And so from our side, there's just the increasing ubiquity of robotics is essentially inevitable. And so when we think about where August Robotics is going to be in one year, five years, 10 years from now, it's going to be leaning into that trend, right? And so, you know, one year from now, maybe we'll have four robotics businesses. In five years from now, we have 40.

1:02:54In 10 years from now, 400, right? Like lots of different robots, all helping people do jobs that they don't want to do. and hopefully many of them collaborating with one another and adding more and more value across the economy and helping people to have more time for things that are important rather than doing these jobs that no one wants to do. Mic drop. Thank you for coming. Very good. Thanks so much.

1:03:20Thank you so much for joining us for another episode of Wild Hearts. If you want to learn more from other ambitious people building, designing and creating the world that we all want to live in, then please hit the subscribe and follow button. It would mean the world to us, the founders, the operators, and the investors who join us on Wild Hearts. This podcast is a labor of love from the Blackbird team and Day One. The show is produced by Camilla Herring and Melia Rayner at Blackbird. Our marketing genius is Eva Telemachus and our editors are from Day One, Annie Jones and Sanjay Chabaria. Thank you all so much for listening.

1:03:59and we'll see you next week.