Dirty Jobs, Real Robots | NZVC Portfolio Day

There will be a minimum of 100 million humanoid robots in people's homes by 2035. AI is going to bring back a resurgence of arranged marriages. I think people are going to trust AI so much more and they're going to sort of outsource all the decisions to AI. Your humanoid robot walks into the lounge and finds a single sock on the ground. What the hell is it going to do with it? The people that will get on board with things and find their place within the new changing ecosystem will be the ones that succeed. And the ones who are starkly against it, I do think they'll be at a loss. The damn construction people keep changing the environment.

So the robot's like, "Oh, there's a new wall today. Now I'm lost and I'm going to sit here and wait for help. Don't just have a pipe inspection." Rob's like, "What kind of pipe? How big's the pipe? What's in the pipe? Is it water? Is it oil? Is it always full? Is it empty? Is it explode? Just go and do it and see what happens. Ideas are worth like jack Absolute jack Like if you can't execute, there's nothing. just take that initiative to actually take whatever crazy ideas that they might have and just try to see what comes with it. Don't be afraid of ambition. Chase it. Love it. Wrap your arms around it and and say it out loud.

New Zealanders don't know any better. They will just get up and give it a shot. That's why we were first to get to the top of Mount Everest. Go to dream that ultimate dream like what is it that you ultimately want to do? And it's all achievable. Yeah, we're going to jump right into um more atoms than bits, I guess. Um and going to start with our robotics panel. Um we have three great speakers um today um who I'm going to ask to introduce themselves, but um when you do that, um can you also give us one prediction about, you know, what what will 2036 look like in New Zealand and beyond? Um and starting with you, David.

David Ings. I did a robotic software startup about 8 years ago, rolled that into a company called Drone Deploy in the US, which just exited to a um New York Stock Exchange company called Procore. My prediction is effectively that there will be a minimum of 100 million humanoid robots in people's homes by 2035. Now, that is an underbaked number and a very conservative number. Uh if you look back at 1977 when I was born, Ken Olsen, the founder of the largest computer company at the time, said there is no utility for computers in people's homes. Microsoft at the time was obviously betting a computer in every home on every desk. Turns out one of them was right.

It took Microsoft 27 years to get to a billion PCs. It took Apple 10 years to get to a billion phones and they didn't even they didn't even invent the phone. So, we're talking about 9 years time. Obviously, Elon Musk thinks there's going to be a billion humanoid robots by 2035. And that's just his robots. So, I'm saying, and there's 130 million US households. So, I'm saying 100 million humanoid robots is underbaked. And people are going to say, "But what about specific robots? You're not going to have a humanoid. It's going to be a cleaning robot and a vacuum your vacuum robot and a, you know, lawn mower." If anyone's ever used one of those and realized, you know, the lawnmowing robot cannot scoop up the dog poo, the cleaning robot cannot clean the lounge before it has to vacuum it.

That's why the humanoid robots will be there. Way more than 100 million by 2036 in the household. In the household. I hope not. I uh I don't like that future, but I I do agree that probably is most likely outcome. So, I'm Hamish. I'm based out in Christ Church. I'm a mechanical design engineer, but I've just started a company called Stanstead. And to keep it short, uh, we use robotics to build timber houses. And we're sort of starting with timber frame and truss and factories, which are then shipped aside for rapid assembly. And so that's kind of a brief overview. I was going to give a prediction for humanoids, but we'll go away. This is one I thought about this morning that I think uh AI is going to bring back a resurgence of arranged marriages.

That's what I was going to say. And uh I could justify it. Are we going to justify it or we pay? I'm so curious. I want to know more. I think people are going to trust AI so much more and they're going to sort of outsource all the decisions to AI so much more. And people are getting quite fatigued by Tinder and such. Uh and I think people are going to trust AI to make this huge decision for themselves. I hope it doesn't happen, but I suspect it will. You can tell it's a robotics podcast when we're talking about arranged marriages. Hi everyone, my name is Sean Donald. One of the co-founders of Purink. We build autonomous robots for pipe inspection.

I guess humanoids has been covered arranged marriages and AI taking over for you know human autonomy. I I think where there's a dangerous or I guess a dirty or industrial job. I think in 2036 we see a lot of that being optimized by robotics. I think we could say replace but I think then you kind of that labor and those skill sets need to move. I know in our world the people on site have such a tacit knowledge for what they do. So it's not about replacing those people and getting them out of the field. It's how do you get them out of harm's way, out of the inconvenient and unscalable, inefficient roles and get them into something more valuable.

So I do see robots by 2036 taking over all of those mundane annoying tasks that really shouldn't Yes. sewage, wastewater, energy, mining where there you convince somebody whose day-to-day life is sewage and pipes to use a robot. It's it's almost the opposite of the way you talk to investors to be honest with you. you need to keep them almost two or three steps back from where you are with the technology because I think especially in today's world they're not you know they see they're on social media they know AI is coming they know AI exists so if you go to them and say you know it's and you throw too much AI slop or you know AI buzzwords at them they freak out and they close up where we found success was really being super practical by there'll be some customers where as soon as you can speak the language know the lingo know basically

the issues they're facing on site and how the solution solves for it you reduce that first barrier and then they're open to talking we've had customers who before we could put a robot on site, they wanted to bring us out on site like physically no solution, no nothing just to make sure we actually understood. So it's like a deep level of domain expertise and understanding that really gets you in in that first opportunity, right? Hey Mish, you're also just starting with uh talking to customers. Um you've built a first prototype. I think when you first talked to Mark, you sent him the video of your prototype and he watched it like 30 times. He was very impressed.

Um, but like how do you go about people that this first prototype is something potentially like saved them tens of thousands of dollars? Yeah, I love showing people that, but yeah, I've literally always got on my phone ready to go usually. So, um, you should ask him later. Ready to go. But yeah, yeah, convincing people to sort of trust your robotics. Uh, that's yeah, one thing I learned really early on was when you're introducing a new idea or new innovation, you want to try and keep it as similar to the current workflow as possible. And so uh one thing I find that with robotics companies coming in is that they they introduce robotics and they they ask the the company or the business or the consumer to completely change their whole workflow around their the new product which might be good and might solve a lot of problems for them.

But people are really stuck in their current workflows. And so one thing I found with doing what I'm doing is being able to offer the same product to these companies which is sort of timber framing but we're able to cut costs a lot and save them a lot of time on site but quite frankly they might not even know it's robotics company but we're able to eventually start introducing that to them uh once we gain their trust and they're more willing to take risks and so that's what I always keep reminding myself just try and stay as so close to their current workflow as possible. I want I want to ask a question to the broader group.

Yeah, the question about humanoids and and you guys are building Himish and you you guys are building very kind of niche solutions, right? Very specific industries. You're becoming experts in that industry. Where like how do you think the future looks? Yeah. In 10 years, is it like humanoids, the general tasks supported by kind of very specific types of robots for for niche tasks or is or will the world's made for people, right? So if you have a human or you in theory can do any task that we currently can do with people or or is it where's a step stepwise process where this first next whatever 7 8 10 years we have to have very specialized robots until you get humanoids to be good enough then you like what how do you think the the yeah the future evolves.

I think like you you said a very interesting thing where the world is built for humans and then a humanoid robot then I guess slots straight into that and that does make sense but like there's a chasm that you know the world and technology will need to cross before humanoids can like one for one replace you know human operation. I know at least in our use case like you wouldn't send a humanoid robot to go you know actually go inside of the pipe. That's just not feasible. It's not going to happen. But there is a world where you may be sending a humanoid out to go and either check up on a robot that's gone through, do a spot check, deploy a robot, whatever it might be.

So then I think it gets to a point where there will be a transition of like you said, human humans are alongside these purpose-built robots and depending on the use case, the purpose-built robot will stay, but then potentially there are grounds to then shift that towards a humanoid. A lot of that depends on conditions and environments and I know for us there's certain deployments in like the energy sector or in the mining sector where it's dangerous for a human that landed on a robot that can combust and all the other things that will come with that. So it's possible I think but there's just a lot more work that needs to happen before it's actually practical.

Yeah, I think it could go either way. I think in in the home I think humanoids have a a real real use case there. They're so adaptable to do so many different things at different stages. I think it makes a lot of sense in factories. I am skeptical. I think it will tend to be more specialized machines. I think maybe in the in the early stages of humanoids kick off, they might in the meantime sort of fill that gap. But I think it will then tend towards a specialized machine because specialized machines are always going to be able to reach a higher throughput, higher accuracy and be more maintainable and have a longer lifespan. We we designed machines that last 20 years meant to run 24/7 uh and you know sub millm accuracy and I just I don't think I could give that same confidence in a in a humanoid not yet anyway.

So just super quick what you said was also really interesting because I think with the context of a warehouse or a logistics center or some factories yes they're currently designed for humans but I say this lightly it probably wouldn't be that hard to just convert them into a completely 100% dark factory where you don't actually need a human inside of it. So then why deploy a humanoid robot into it? So I said I think it it will be very much use case based. Yeah, I think my background was in software and we always talk about designing software for what they call the happy path. The happy path is where a customer does exactly what you expect and want them to do.

And of course the testers on the software side know that humans are slightly unpredictable but software is really good at forcing users to go down a certain path. When you get to the physical world you can't control those exceptions. And so where robotics is thriving at the moment is where you can control those exceptions in a factory environment. So Amazon didn't decide to make a humanoid robot that can pick something off a shelf and put it in a bag. They designed their physical environment to build the robots and they changed the environment to match the robot environment. So they can build a happy path and completely control the environment. You can then extend that out to agriculture.

Farmers for thousands of years have reconfigured their physical environment or the machines. Bigger wheat production, open areas, you know, the apple picking robots nowadays, they don't say, "Hey, can I get that robot to work in the current apple orchard?" It's how can I rebuild my apple orchard to work with these robots? That's a great example where the humanoid robot's not that useful. So, if you take environments which you cannot control, um, a household is one. So picking stuff off the floor is is a really big problem. Driverless vehicles, the reason it's taken so long is because these poor cars have to survive an environment where you can't change the physical environment. So I think what we say in robotics, the biggest problem with hardware and robotics is exception handling.

And so the humanoid robots and these are going to play a really important part for the exceptions. And the question is how long do those exceptions still exist is is the big challenge. David, do do you see any parallels with how software developed over the last I guess whatever 30 years where first you buy software on on discs, right? And then eventually you had kind of the internet allows you to, you know, maybe go visit a website and then at some point bandwidth constraints allows you to download stuff directly. I mean, and then you had mobile like so sort of it feels like you kind of kind of, you know, saying there's a lot of different form factors, a lot of different use cases.

They all kind of require a different approach like a do is that is that true? Do you see some parallels there to how maybe software rolled out over over the last 20 30 years? And also then did do you have any predictions for what that looks like? Is is it is it a fool's errand to build humanoids now? May maybe right now is the time to build very specialized or or maybe the opposite. I think the good news is there's enough money in the ecosystem that is building all robots for all utilities that they could possibly imagine right now. And so that evolution will determine where it's successful. I think there'll be an unpredictable nature of I keep saying, okay, the lawnmowing robot maybe is a bit of a pain because there's all these exceptions, but the reality is you can also outsource that where somebody comes with their own lawn mower and does do

all of those exceptions for you. So, so maybe your humanoid doesn't go outside, but maybe it's not uh the fact that you have a lawnmowing robot, it's the fact that something arrives at your house, your lawns magically mowed, and then they vanish, and you didn't have to worry about it. So that's the what are these evolutions of as a service that start existing versus you know it's not practical to not have a washing machine yourself. It's not practical to not have a dishwasher yourself and you're never going to get your humanoid robot to wash your clothes manually. That seems like a really bad idea. So that's an example where you're going to have a specialized machine that does the job.

So I'm I'm just pretty excited to see what all those models are going to be. And I think the concept of buying a humanoid robot that's probably only going to be out of date within 2 or 3 years, that doesn't seem like a really practical model. It's much more going to be uh for police or a utility. Um but but just like driverless vehicles took so long, I have no doubt that humanoid robots are going to take a long time as well. And I always use the sock example. Your humanoid robot walks into the lounge and finds a single sock on the ground. What the hell is it going to do with it? Does it go in the washing basket?

Does it fold the smelly sock and shove it in the drawer? Uh, does it find the partner to the sock? There's just a million examples like that that this poor humanoid robot's going to have to have some default for. So, it's still going to feel pretty dumb for a very long time, right? Yeah. Interesting. Um, maybe to pivot a little bit on the on the commercial side. I think in general, but robotics specifically, people get scared, right? It's like I'm going to take my job is going to take whole industries out. How do you guys and and then when you talk to customers I assume for a lot of the customers you talk to like you're kind of replacing a large part of what they do in some in some instances right like like amish you're actually looking to compete with some of these frame builder and then at the limit the limit you ideally

want to just do the the entire industry yourself right you just want to put everyone out of business but you can't and maybe that actually happens maybe it doesn't obviously you don't want to say that you want to be genuine when you when you say that you don't want to do that right you want to build trust with the how do you think about that how do you manage that and and what is How do you find the wedge into commercializing building trust with people and and companies who you may eventually replace? Yeah, this is this is a tricky one because you know I personally get a lot of fulfillment from my job and the idea that like my job being automated scares scares me you know but I I think uh what gives me sort of some reassurance is that especially in construction like there's just such a shortage of laborers.

You go talk to any builder and um you ask them what their biggest problem is. was finding good quality labor. And so I I do think there'll be jobs that sort of relocate, you know, some so I might put some people out of jobs in in a permanent trust factory, but there is definitely work for them on site and so it might sort of, you know, cause this sort of shifting. So that does sort of give me a little bit of hope and you can sort of go into the jeans paradox about these things but it is just sort of like if we can make housing so much cheaper then that's just going to hugely increase the demand of houses and more houses will be built and you know the cycle continues.

So that's why I try to help people. It's pretty easy for me to sleep at night, which is nice because I think we're one of the unique use cases of robots where like the human actually doesn't want to be there. Whether it be an open sewer, whether it be like a spouting water man any like on an offshore rig, no one actually wants to be there doing the actual thing that we're solving for, which is fantastic. So we don't actually get a lot of friction on the physical doing side where I think and this is only some customers who are like actually if we've really integrated with AI and like think about the future is like when purink has robots living inside of pipe networks and we've got like a constant kind of like managed service running but I think they also understand that their knowledge of the industry isn't going to be wasted.

So whether that be them joining purink or whether they are the ones who are maintaining the pur link robots while they're out there there is a place for everyone you're getting to. The other thing is I haven't spoken to any customer in the last 18 months who's like, "Oh yeah, I've got too much work. I don't have enough work to do. They've got so much going on." I'd assume, especially with you guys, right, with housing crisis that we're in globally, but also if you can make that more efficient and actually help people do more. I think people are here for it and they kind of understand that. I think the people that will get on board with things and find their place within the new changing ecosystem will be the ones that succeed and the ones who are starkly against it maybe in the short period of time will maintain whatever sense of normal they

have but eventually I do think they'll be at a loss but in the long term I think is where the the drastic balance will come but for the time being I think people are really happy with the efficiencies that they're seeing. Yeah, we got a bunch of questions um from the audience. One is on 3D printing. So, any developments coming on 3D printing that may enable more local and bespoke construction of robotics? I've seen it until your office. Yeah, we've got three now. And then we've also got a five axis CNC. So, we're we're moving away. So, we actually looked at this as well cuz with, you know, the recent expansion for Purink, we were like, "Okay, we need a way to quickly prototype." We looked at a metal 3D printer, but right now with, you know, the amount of effort that goes into one putting the printer on and then all the post-processing you

have to do, it's not the same as like, you know, a bamboo or like a, you know, filament based 3D printer, that'll probably change at some point. But for us, 3D printing is like the first step. Unfortunately, there's some parts that you just cannot 3D print and put inside of our robot because of how small it is and the precision we need. But on a macro scale, if a metal part breaks, there's a few that we can replace, get out there, test it, but not something we can put out in production to the customer. Yes. Yeah. In construction, there is a few times that, you know, you've obviously seen there's like a whole convention of people trying to 3D print houses now.

It's nuts. It's really kicking off. Uh and I and I I really think it's people who have just sort of use a 3D printer on the desk and being like, why can't we just, you know, build a house like this? And they sort of get carried away with the idea and they don't really sort of go talk to customers about it. And then you go talk to builders and they're like I am not touching this. What the hell? Like it's just again it's such a different workflow for everyone apart in in the construction process. So the architect has to design a new house. The consent has to consent this new material and this new construction engineers need to calculate this new material and then sparkies plasterers.

Everyone on site has to learn a new trade and how to construct it. And so I personally don't from the data it doesn't actually look like it's any cheaper and then there's this added sort of barrier of just it's just so hard to get builders to trust it and try it. So that it's not a feasible technology in my opinion. be fair though what I found especially with our industries if you find not the weakest link isn't the right term but if you find the specific role in the value chain that's easiest to like the either the most under the pump or the most cost constrained for us it's the people doing surveys inside of sewer networks right it's such a low margin business but they have to do it because if they win that work then the first one to find the defects and then they do the high margin uh remedial works right

so when we go to them and pitch them an autonomous solution that can help them find the defects faster with less people on site all that sort of stuff they're the first ones like yes let's do it Right? And then all of a sudden you find the people who are doing the repairs and all the other parts and all the other people in the value chain start to see other people benefiting from autonomous uh robotics and then people go, "Oh, why don't we have that?" And then you kind of see it. So it does like you're saying, it's a hard slog at the start, but then once everyone starts to see it and everyone starts to want to get a piece of it, then after that you just have to be ready for the for the influx.

Yeah. Um, one question for David or I guess about David's point on designing robots for environments or designing environments for robots. Which which one of those is more likely in the future? So, are we going to design robots for the existing environments or are we going to envi design environments for robots that we're building? Yeah, I think both both are obviously true. I think if you take again autonomous vehicles as a great example is at what point will we start changing our road system for the autonomous vehicles and people say oh that's absurd we'll never do that it was like we're already changing our road system for electric vehicles in some way we're putting in charging stations so it's not unfeasible obviously those cars are pretty good at reading physical signs but then also they have all the digital data on what the speed is and things like that so so I think it depends

on the adoption rate I think a lot of the time So where we can uh the problem is who also owns that infrastructure does a government have to make the change so that Tesla can sell more autonomous vehicles or can the people selling the robots actually get involved in that physical environment. So you know or an owner owner of a farm and yeah so often it's going to be an ROI calculation with government infrastructure. there's going to be politics involved in terms of long-term commitments, but I think absolutely the environment will be the first thing that changes because usually the ROI is much greater in changing the physical environment. Factories, farms as I said happen at mass scale but then other environments where which are very difficult to Yeah.

Yeah. that wouldn't be but the question is in the home we've seen even over the last 10 years I've seen many formats of kitchen robots that can make you know food and the problem is the amount of physical augmentation you need for a kitchen to be viable even in a commercial kitchen it's almost like you have to lock the whole thing off everything has to be in the perfect place and then the pitch is ta-da we can make soup you know it's like so so I think that's an example where you you you you will see some changes in the kitchen maybe for the humanoid robot to make it easier to cook in a kitchen, but you're not necessarily going to see people that isolate their kitchen off to just have a cooking robot, you know, on the end of an arm.

Mind if I jump in. David, you you're you sold your company actually sold it twice, right, this very last time to to a software company, Procore. They do they do project management for construction sites. So, why did a software company buy kind of your Yeah. your the robotics company? Yeah. So the the premise of Rokos, the company I co-founded, was really around operationalization of robots. So we didn't build a robot. We were saying there were lots of people in fantastic universities that were building robots. Those people, it turns out, were more mechatronics, electronics engineers, and they knew how to build physical things to do physical things. There's a hell of a lot of software involved in that process, but they're not cloud architects.

They don't know how to build cloud things in an operational way. So the question was great, we've now built a robot that can pick strawberries, but to your point, how do you operate that robot? So it's the do you have a PhD from Stanford following the robot around every time there's an exception or do you need to connect to it remotely and be able to give it a mission? Hey, here's the strawberry patch. Tell me when you know you're done. So the premise was we built software for automating other people's robots. And that's again the ecosystem play. So we assumed robots would have APIs. there would be a device that you could control from a remote location.

Uh so the the the the Procore acquisition and originally the drone deploy acquisition was in these construction environments and in operation and maintenance, you increasingly want to get what they call reality capture. So you want to take photos of the wall before the drywall goes up so that when something goes wrong, you know exactly what's behind that wall or you can verify that the contractor has done that job. But it's very time consuming and actually looks ridiculous where we call it the unicorn. It's a hard hat with a 360 camera on it and the person's walking around the site basically just capturing data. But if you don't capture that data, you can miss things that the subcontractors didn't do.

You don't have insurance backup which is say, "Hey, it was done properly, but now it landed up bad." So, so the real rollup that's happening now is these robots in the air and on the ground are capturing these sites on a daily basis and capturing that that physical version of reality. It's like a database backup of of reality in the sites. So the software side of things is controlling that. But increasingly, as you said, hardware is getting easier. So So one of the services that that at drone deploy and procore we now offer is that scanning as a service. So the robot still has these weird exceptions, but instead of calling their the general contractor superintendent, hey, please find the robot.

It got lost and got out of Wi-Fi coverage. we just deal with all of that so that they're basically just paying for a daily scan of their entire site into our software. And at some point the question is maybe the exceptions come right down and you know there's only once a year where somebody has to go find the poor robot that got lost and didn't manage to make it back to its charge station. Maybe at that point we unlock it and say hey now you can just lease the robot. But until that point we need to deal with Wi-Fi and and all of all of those issues. So, so the roll up firstly is really how does uh Procore as a software company help their capture capture the data that becomes 10 times more valuable because if they don't capture the data our software is not as valuable.

So we kind of taken care of the software capture side but the next question is how far into that hardware world do we go? Do we rely on everyone else building great hardware? So one of the premises at ROCOS was DJI did a phenomenal job for good or bad just decimated the drone industry. They were 10 times cheaper and 10 times better. So, at the moment, that's the challenge with the DJI Ice. It is a phenomenal drone. It is just way cheaper and way better than anything else that's possibly available. We thought the same evolution would take place in rovers and quadripeds and these different form factors. And it really hasn't. They're expensive. They pretty useless.

And a lot of the time, the problem is just exceptions. In the air, it's actually way easier to get from point A to point B. You only need three or four form factors. How big's the payload? How far does it need to fly? How long's it got to be in the air? On the ground, it turns out there's stairs and there's mud and there's rain and there's getting lost. And the problem in construction is robots are good at mapping something, but the damn construction people keep changing the environment. So the robot's like, "Oh, there's a new wall today. Now I'm lost. I'm going to sit here and wait for help." So, so, so it turns out that was wrong.

It wasn't. Those robotics companies weren't going to get there. So then we said, "Okay, well maybe we need to build our own so that we can control that environment." So, so yeah, we're a software company, but we having to dabble in hardware a lot of the time. I feel like a lot of the industries now, especially in the context of robotics, are in that weird teenage pimply phase. They don't know how they feel about themselves. They're going through it and then there'll be this maturity that will come in whatever time scale. I don't know where this all closing comes together because of the awkward and weird scenarios that robot has seen over the last x amount of time.

And that's I think where that high level automation does actually end up coming in. Yeah. Yeah. And I think for you guys, what we've seen even in that contained vessel environment is you don't just have a pipe inspection robot. It's like, what kind of pipe? How big's the pipe? What's in the pipe? Is it water? Is it oil? Is it always full? Is it empty? Is it explode? You know, can it explode? It's just there's so much variation that it's very unlikely to be one company to create, you know, one robot that can do all of those things. It's just not viable, you know. So, you have to pick your use cases. Yeah. Yeah. Maybe one last question from the audience.

Um um and that's probably an interesting one for for New Zealand and spec specifically, but also for Australia. How much of um what you do relies on China in terms of critical minerals, batteries, parts that you're sourcing for China? And do you see that I mean if if you have conversations in the US, is that any is that a problem? Probably too much right now. Like I think we've we've got versions of the robot that out of necessity we've produced domestically. It's far more expensive and takes a lot more time and a lot more massaging to like get what you need. So China's a lot of times their quality isn't even that great, but they'll just say yes to all of the designs that you give them and you know they'll give you a batch that's off and then they'll send you another batch and they'll keep sending you batches for free until they get

it right. And that's really what folks in the US especially with some of the new legislation that's coming out. We've had to look around our supply chain and find okay immediately maybe we're not going to transition to a fully sovereign whether it be Australia, New Zealand or somewhere in the US supply chain but does it exist and it does but it's just more expensive but doesn't mean that it's impossible. We've even like I mentioned we've invested in a fiveaxis CNC and some other manual machining capability mainly for rapid turnarounds for R&D and small production stuff for customers. I think robotics companies at a certain size, that's where they'll also start to see benefits is if they can do things inhouse themselves without it being ridiculously expensive.

So, so, so you're seeing it more as a speed boost, right? You don't have any kind of national defense. I don't for you it doesn't really matter what you're making it, right? Yes and no. I think it also depends on who you talk to in the like for us, right? Depending on the region, the utility is a government-owned organization. We normally work with the contractor, so it's a private entity, but then depending on who who hears about it, they ask for different things. So there's definitely a lens of it will become important in probably the medium term for us, but not something I have to deal with and keeps me up at night in the short term.

So basically, if China hacks your robots, they can hack the utilities basically, right? That's I I wouldn't go as far to say that, but it's one of those things where if we were fully, you know, a lot of our stuff is custom. I think the really only offtheshelf component is the Nvidia chips we use. So if they can hack the Nvidia chips, Nvidia's got a problem, we've got a problem, everyone's got a problem. But yeah, like to an extent, if we're fully integrated and we don't have a sovereign supply chain of some sort, that is a concern for Stanstead, we like for the for the prototype we've built, I think about 95% of the parts came from China.

And you know, if that wasn't an option, we had to go to Germany for for servos, the whole machine could have cost 30 times more. like it is it is uncomparable in terms of how expensive it is. And so if if if we sort of lost the ability to to order from China, that would just almost completely halt the the the business case of standard because right now we can build machines really cheaply and it's sort of because the cost of automation has come down so much because China's just like commoditizing all these very complex components. Uh, and so that is, you know, a huge factor when when designing machinery, especially for prototyping, which is now really much more easy to do.

In terms of timber, it's really crazy. I like to tell people that we send our logs to China. They then turn them into engineered timber, sell them back to us, and it's 60 cents on the dollar if we were to to get them manufactured in New Zealand. That's crazy. We need to fix that. Um, we need to try and capture some of the value in in New Zealand. And that's that's one challenge which we're trying to address and so it's we'll probably take a hit in the in the early stages but I think trying to build that capability domestically and especially in forestry I think it's a worthwhile mission today. You see it in metals with mining as well with Australia.

It's you look around and look at the exports and then you'll see what's coming back in and you scratching your head why is this things have happened and that's just how it is. But I I do think there needs to be a shift back and like you said there will be hits. Things will cost more. It'll be expensive, but I think over time it will start to even out, but it's a worthwhile investment for the future of both Australia and New Zealand. Someone said something very interesting. The reason Australia and New Zealand typically come out with such like cool innovations is because we have a good quality of life and we generally take care of our workers, which is fantastic.

But that also means labor is expensive. So, we need to find ways to be more efficient and produce more without having to hire more people. And that's where you see cool automations and robotics coming out of Australia. I think more of that needs to happen on the manufacturing front and there also needs to be a focus of bringing manufacturing back to all good. Great. I think it's a great first panel. Thank you so much for joining us. Thank you for tuning in to Techmates. If you enjoyed this episode, be sure to subscribe, leave a review, and share with friends. We'll be back soon with more stories of Kiwi and Aussie founders reshaping the future and disrupting Down Under.

Until next time, keep dreaming big and daring to disrupt.