YC Tech Talks: Climate Tech with Charge Robotics (S21), Wright Electric (W17) and Impossible Mining

[Music] I'm Paige Amora. I work at Y Combinator. I'm on our work at a startup team, so we're the team that helps our portfolio companies hire. For this event, we'll do three tech talks. These will just be about a technical topic that the founders find interesting or a bit about their company as well. Then we will have two quick pitches to hear about other companies, and then we'll initiate breakout rooms. So with that, let's go to Jeff.

I'm going to talk for 5-10 minutes or so about the electric airplane industry and in particular energy storage. I'm going to cover a few different areas. First, a little bit about the aerospace industry and the particular market segments that we're working on. Number two, some of the main technology challenges, and then some of the major opportunities in the energy storage space, especially funding opportunities, which I think will be relevant for early-stage startup companies. I'm happy to take questions at the end.

Just briefly about us, we went through YC in 2017. We're about 20-25 people. We're based up in Albany, New York. We're funded in addition to private investors by NASA, the U.S. Department of Energy, the U.S. Air Force, and the U.S. Army as well. Very briefly, the carbon footprint in the aerospace industry is very, very bad. You would have to eat vegetarian for over a year to offset the carbon from even a single relatively short round-trip flight.

We focus on electric solutions to the aerospace industry because there's a bunch of research that's come out that other different technologies involving combustion—things like biofuels or electrofuels or even hydrogen combustion—when stacked up against jet fuel, aren't even necessarily that much better. Electric technologies have the potential to be substantially lower in total greenhouse gas emissions.

So what this chart looks at is not only carbon but it looks at carbon nitrogen oxides, contrails, and a bunch of other things as well, and tries to get to a total global warming equivalent. What you see is that in some cases combusting hydrogen is substantially worse, and electric is sort of in many ways the best environmental solution.

When we looked at the aerospace industry, we were surprised to see that when companies like Boeing announced a new airplane that has, let's say, a 15% improvement in fuel burn compared to a previous airplane, a lot of the underlying benefits come from engines. For example, the 787, which had a 15% fuel burn benefit over the 767, a lot of that came from propulsion technologies.

We then looked at the aerospace industry and thought, "Wow, there's a lot of people working on relatively small airplanes." But unfortunately, small airplanes smaller than 100 passengers represent less than five percent of the carbon footprint of the entire aerospace industry. In fact, most of the carbon footprint of the aerospace industry is airplanes larger than 100 passengers, and in particular in what's called the narrow body segment.

So that's the 737s, the A320s. A typical airplane, like you might be on, carries sort of 100 to 200 passengers, and so that's the industry that we work in. For the business development people on this call, this is an enormous market. These are companies that are each worth 100 billion dollars. There are expected to be 30,000 new narrow body airplanes purchased over the next 20 years—billions and billions of dollars of jet engine sales per year and billions of gallons of jet fuel. So this is a large potential market opportunity if you could create the underlying technology.

And so what's needed? Number one, propulsion, and number two, energy storage. Propulsion essentially being ultra-lightweight engines; energy storage of ultra-lightweight ways to store electricity. So this is our electric engine that we built. It's basically meant to be a replacement of a jet engine of a narrow body class airplane. It's a couple megawatts in power, which is about 3,000 horsepower, substantially more powerful than anything that's out there in the aerospace industry.

In-house developed inverter designed for high altitude operations. Just finished some testing at the FAA technology center and planning for FAA Part 33 certification in the 2025-2026 time frame. So that's kind of a little bit about what our company does.

What we're planning to do as our next step is we're taking a 100 passenger airplane and we're removing one of the four engines and replacing it with our electric engine to make it essentially an electric test bed. So this is one sort of quick point to the group: if anybody is working on propulsion technologies and is looking for a relatively inexpensive way to do testing, please reach out to us because we're already going to have this platform in place. We're happy to do testing with you if you're working on technologies like this.

The nice thing about this airplane is it's a four-engine airplane but it can actually fly on three of the four engines. So we're using basically one of the four engines as a laboratory in the sky. I want to talk now a little bit about energy storage requirements. So no lithium-ion battery is going to get to, you know, greater than 750 watt-hours per kilogram. Typically, lithium-ion battery best case scenario is going to get up to 400, 450, 500, sort of absolute best-case scenario. So you really need to be looking at new chemistries.

What this industry is really looking for is even above a thousand watt-hours or 1,500. You also need high C rates; you need relatively high discharge. It doesn't have to be as high as the vertical takeoff and landing airplanes, but you know, sort of three to five Cs. It also has to be altitude capable. In terms of energy storage, what you really need is electricity. You don't necessarily care if it's a battery.

So the two different technologies that have been mostly in development are sort of energy carrier technologies. One quick note: if you look on the right, this is essentially the energy profile. You have a huge amount of power up front during takeoff and climb out, and then you have about a third of the power needed during cruise. This is a similar path; you see a huge spike in power needed right here, and then it sort of drops off for the power that's needed—it's about a third to a quarter of the power needed.

So two different major technologies under development, and this is what we would encourage people in this group to work on. I'm very happy to talk about this with people later. Number one, aluminum air fuel cells or other metal air chemistries, and number two, hydrogen fuel cells.

So in terms of aluminum air, aluminum air is generally in the category of other metal air chemistries. Zinc air, for example, has been used in hearing aid batteries since the 1970s. Forum Energy is a company that's raised about $300 million, maybe even more than that, to focus on an iron air battery. It's a new chemistry that’s sort of something old but new; a lot of people are working in this space.

Actually, at the most recent U.S. Department of Energy RPE summit earlier this year, there was specifically a group that was talking about putting together funding related to this topic. The disadvantage of this is that you can't recharge these batteries; you have to recharge them by taking them, you can't recharge them by plugging them in. You have to recharge them by sending them back to a facility.

So there's a lot of operational challenges with this, but at the same time, it's a major advantage. I just have two more slides. A lot of people are also working on hydrogen fuel cells. Obviously, everybody knows what a hydrogen fuel cell is, but there's a lot of applications potentially in the aerospace industry, and people are looking at technologies that could be greater than a power density of two kilowatts per kilogram by 2030, and that’s sort of what a lot of people think of as the number that's needed for large aerospace applications.

So if you're working, for example, on ultra-lightweight fuel cells, aerospace is a great industry for you to be in. The last thing I want to say in this space is there's a lot of money going into the space right now. So, you know, everyone wants to go get venture capital money. As I mentioned, Forum Energy has raised a whole bunch of money in this space, but here's just four different opportunities or three different opportunities that are sort of live opportunities to raise money. If you're in Europe, there's the Clean Skies; I think they've put up 700+ million euros working in this space.

ATI Fly Zero is a program out of the United Kingdom. The U.S. Department of Energy is doing a lot of work, and then there's obviously other things as well. So we think this is a strong space because it's a huge potential market, there's a lot of money going into the space, and major opportunities to revolutionize the U.S. aerospace industry.

So I’ll pause there. I wanted to keep this speech relatively short. Please feel free to reach out to me. Take down the email address. I'm not going to be able to stay for a breakout session, but if folks want to schedule one-on-one call, I'm happy to do so. So thanks very much for the opportunity!

You have a couple of questions here in the chat. How were your relationships with the DOE in the Air Force established? Oh, great question! You know, so let’s see. We've had a few different contracts with NASA, a few different contracts with the Air Force and with the Army. Some of them are sort of serendipity; NASA puts out a request for a proposal and it’s exactly the thing that you’re working on. That’s a sort of luck, but you know, it occasionally does happen.

But sometimes also, these organizations have open solicitations. For example, with the Air Force, we’re being funded now to turn on to an ultra-lightweight generator for them, and that came about because we found a unit within the Air Force that was looking for an ultra-lightweight generator, and we said, "Hey, we could do that for you!" And then we sort of put together a proposal.

So I would say it's, um, maybe the short answer is, some advice that I was given during the YC batch: try everything you can think of, and then try it again, and then try it again, and try it again. It’d be nice to say, like, "Oh, it’s just an easy path," but really I’m just sort of trying over and over.

Let’s see, what do I think of the ion engines which made flight a couple years back? I don't know as much about ion engines, but if you send me some information, I'd be happy to talk with you about it. Are these the ones that came out of MIT? They’re, uh, I think this is the concept that it's an essentially no-moving air concept. If that’s what you guys are talking about, I saw the video too, and it’s like absolutely insane.

It looks like it's relatively small, but, you know, obviously there should be tons of money going into that space. Next comment: it could be cool to use a catapult. I agree 100%. Catapults or ramps or other things I think have a big opportunity. It's just that’s a lot of infrastructure on the ground, but no, I love that concept as well.

Yeah, so that’s exactly the concept. You were talking about how does an engine compare to engines used for smaller planes like that were used for firefighting? Is it adaptable to that purpose? In fact, it is actually. The airplane that we're working on is a 100 passenger airplane called the BAE-146. So, I’ll go back to it. It’s actually used in the United States as a firefighting aircraft, so it is adaptable for that purpose.

I think that’s the last question. So thanks, everybody. Next up we have Max from Charge Robotics.

Hi everybody! First, I just want to say thanks to Paige for setting up this event. My name is Max Justice, co-founder and CTO of Charge Robotics. We were in the Y Combinator Summer 2021 batch, and Charge Robotics builds robots that build large-scale solar farms. When I talk about large-scale solar farms, I'm talking about like what you see in the background of this slide here.

So these are pretty huge sites; they're often hundreds or thousands of acres large, generating hundreds of megawatts to sometimes gigawatts of power. So a couple quick things to know just about the solar industry in the U.S. right now: about four percent of our power in the U.S. comes from solar, and by 2050 it’s going to be about half.

So there is a tremendous amount of solar that needs to get built in the next few years. The problem is that labor is actually the key bottleneck preventing solar adoption right now. So if you think about it, it kind of makes sense: these sites are built often out in the middle of, like, very rural areas, and getting hundreds of people out there for months or, you know, sometimes years at a time is just really challenging.

So, you know, as Charge Robotics, like I don’t really know how to source 300 people to get them out to a solar site for a year, but I absolutely know how to ship a couple of shipping containers full of robotic equipment out there to build the site. So that’s kind of our thesis: we think that robots are the right solution to solving this labor shortage.

If we're talking about deploying robots to build solar farms, I think it's important to kind of understand the sorts of tasks that these robots are going to be doing. So I want to talk about how solar is actually built today. There are kind of three main mechanical steps that go into basically every solar farm that gets built. First is pile driving—so you send out a crew and they hammer these large metal I-beams into the ground in a huge grid that covers the whole site.

Then you send out another crew and they bolt tens of thousands of these large metal rails between the I-beams, and those are actually what support the solar modules. Then you send out another crew, and they bolt the hundreds of thousands of solar modules onto that racking structure.

So those three steps are common to, again, basically every site that gets built today. And I think one thing just to note here that’s kind of funny is that basically everything on a solar site is really heavy. So like, a single solar module, like one of those rectangles, weighs like 50 or 60 pounds. As a single person, like you really don’t want to have to carry one very far.

And so what they do on these sites is called material staging. So they send out crews with forklifts, and they basically move pallets of materials kind of like close to where they need to go. And the idea is that, you know, if I need to pick up a solar module, it’s maybe only 10 feet from where it needs to end up.

And so what this made us realize is that kind of no matter how we’re automating the solar industry, a key primitive that we’re gonna end up needing is this, like, stuff mover—this robot that can move things from point A to point B on the solar site. And so that’s actually what we built towards the end of last year, beginning of this year, is a giant robotic forklift, and that’s kind of what I want to talk to you guys about today.

One second, I think I just realized that I didn’t have it in video clip sharing mode, so it might be a little choppy. All right, I’m just going to unshare and be sure really quickly. All right, that should be coming through a little smoother now.

So yeah, this is how we built that robot, our robotic 25,000-pound forklift. So the vehicle of choice for this application turns out to be this thing called the telehandler. It’s basically a big forklift; it has these massive tires. The kind of relevant thing to know about these is that they’re already used on solar sites today. These are super common vehicles; they navigate crazy terrain.

The construction workers on these sites kind of know how to operate with these already, and if you're gonna automate one of these vehicles, there's a whole bunch of different systems that you have to control from software. So you have the fork angle, as well as the boom extension angle; you need to be able to move those just so that you can pick up pallets.

There's also steering, throttle, and brake as well as gear shifting because you need to be able to go from forward to reverse. This is kind of like the bare minimum set of stuff that you have to control from software to be able to build this vehicle.

The first system that I want to talk about is the hydraulic system in the vehicle. There were two systems that we needed to control that were on the hydraulic circuit here: that's the steering and the brake. So for the steering, this is kind of the sort of complicated looking hydraulic circuit for this steering system in the top left here.

We ended up deciding just in the interest of time to basically bypass that entirely and 3D print a custom bracket to bolt to the steering column. So this actually ended up working great! We ended up 3D printing a bunch of different pieces and then just buying an off-the-shelf motor and motor controller and gearbox to control the steering.

So this is kind of what we ended up with; this is actually what we shipped to our first pilot deployment. The other hydraulic system was the brake. So similar to the steering, we ended up mechanically actuating this, so we had a steel cable going from the brake pedal to a linear actuator sitting underneath the vehicle.

What is sort of nice about this design is that if I'm a person sitting in the cab of the vehicle, I can jam on the brake, and this cable will just go slack, and it’ll still function like a normal brake pedal. And so this was actually a design philosophy that we decided to use for all the systems in the vehicle that we could at least.

So for example, if you move the joystick as a person in the cab, that manual input will actually override any automated control of the vehicle. So this is kind of a philosophy to automating vehicles that we tried to use everywhere. The next systems that we automated that I wanted to talk about today—sorry to interrupt.

I think there’s a—I don’t know if it’s a Zoom window or something, but it’s blocking part of your slide. Is it on the right-hand side? That guy? Yeah, you’re right on it now. Oh, okay, yeah, weird—that's the thing with all the faces in it. I’ll just minimize that; that should be better.

So yeah, so next I'll talk about the electronic systems. So that’s the joystick, the throttle, and the gear shifter. So the first step in getting these systems working was to reverse engineer this thing called the CAN bus. This is a protocol that's common to basically every vehicle that you've ever been in; it’s used to send data from system to system inside the vehicle.

And so if you want to take over control of some system inside the vehicle, you can figure out what CAN messages it's sending and then send those messages for yourself. So the trick then becomes how do you figure out what the relevant CAN message is?

And so what you can do is basically take all the CAN messages and graph them, and then if you move the relevant input or change some variable and you see, for example, a line on this graph shoot up, then you know that it’s very likely you found the relevant message, and you can start sending that message for yourself. For the signals that weren't on the CAN bus, we built this custom cable that sort of just kind of went in between the main computer inside the cab and the rest of the vehicle, and then we could just splice into signals directly.

And so when you tie all that work together, you end up with a vehicle that you can control completely from software. So this was pretty shortly after we got the vehicle. Do not try this at home. This was when we got the boom working on the canvas.

So we quickly kind of whipped together this interface that I had on my phone, and then, yeah, I was able to, you know, take the telehandler for a ride through remote control. Actually, when we were in Y Combinator, a demo that we put on was we sent that webpage to one of our batchmates who was in Mexico, and she was able to control the telehandler from another country, which I thought was kind of funny.

And there you can see, kind of, when we got the joystick working, we had it ripped out, and everything's running off my laptop there. So this is what the early days of Charge Robotics look like. So yeah, then we wrote a full autonomy stack and got this thing deployed onto an actual solar site in Iowa back in March.

So I'll skip through this video just because it's a little bit long, but we dock with an actual palette of solar modules, we do some path planning, you can see the vehicle following a path here, and it's localizing off of some cameras that are mounted to the vehicle. We drive to the final location, and we put it down.

So yeah, this was back in March, and with that deployment behind us, we started moving on to our next major project, which is actually building this portable robotic factory. Let’s see if I can advance. Yeah, so portable robotic factory—basically, there's a bunch of solar hardware that needs to get assembled on site, and so we're building this factory that fits in the form factor of a shipping container that we send out to the site, and it produces those assemblies.

So these are massive assemblies; like I mentioned, everything on the solar site is heavy. These are 36-foot long panels, they weigh several hundred pounds. And yeah, it's basically like a car factory like you'd see, you know, like an automated Tesla factory or something, but made portable such that we can ship it to a solar site.

So here's one of the arms that we got in the mail a couple weeks ago. So with that in mind, if this is the sort of work that sounds exciting to you, we're currently hiring for mechanical engineers, so I’d love if you could reach out if this is the sort of thing you want to be working on.

But thanks very much! Thanks, Max. We have some questions for you in the chat. So Joshua's asking, does the elevation or two terrain surface—especially irregularities or slopes—limit your deployments? Got it. So, yeah, for the most part, these sites are quite flat, just sort of by design. It adds a lot of cost to the site if they have to do a lot of grading.

So like, if the land's not flat enough, they actually send out construction equipment to flatten the land. So it's very well specced how flat these sites are in general, and for the most part, we are focused on the biggest, flattest sites.

What type of control theory are you using? PID, baby! PID all day, every day. That’s what we use to get this prototype working. As we kind of advance our technology, we’re almost certainly gonna be using some more advanced stuff, but for the purpose of that prototype, it was all PID, and then it was pure pursuit was the path-following algorithm because we had an Ackerman steering vehicle.

Cool! How much time in labor does Charge save a typical solar project site? Oh sorry, how much time in labor does Charge save a typical solar project site? Also curious who your first customers would be and why. Thanks, good luck!

Yeah, so it turns out that labor is something like 30% of the project costs, and of that, mechanical installation is about half of that. So what’s kind of crazy is that there's this labor shortage is big enough right now that we don’t really need to charge less than human contractors charge for people to be really interested in our product. So for example, we signed some LOIs with some of the largest solar construction companies in the country for about $31 million, and those contracts were actually priced at the same price of human labor.

So people are just desperate for more installation labor periods, so we don’t actually need to charge less than people. How many fewer people are needed out of deployment site by using a robotic forklift? Some of the videos seem to show a few workers watching the forklift.

Yeah, I was one of those people in that video. This was very much like a prototype deployment. This thing is not ready to, you know, it's not a product yet. Our complete system will require dramatically fewer people than conventional labor today.

So a single one of our systems can do the work of something like several dozen people. I hesitate to give an actual number, but our models have it as something—I think it's about between 70 and 100 people, and that’s for a complete system.

So that’s like six of the robots that you saw. How are you controlling the forklift's exact positioning? Yeah, we're going to be using RTK GPS. We weren’t for that demo, but all these sites have an RTK GPS base station on them, so that’s what we’ll be using to aid our localization.

How long does the forklift last per charge? How long does it take to charge fully? Fun fact: they burned thousands of gallons of diesel per week on solar sites! We did the math to make sure that these sites offset themselves relatively quickly, but every vehicle that you see on these sites is diesel-powered, including the one that we automated.

What is the most labor-intensive part of the entire project? You hear varying opinions on this, but I would say the most popular opinion and the one that I share is that it’s the module installation step. That takes a lot of work. It’s just those things are really heavy, and you have like, you know, hundreds of thousands of them, so I’d say that’s the most labor-intensive part.

If there is that much of a shortage, can't you charge more? Potentially! It’s a great point, very very possibly, yeah. And also, there are things that you can do with robotic installation that it’s easier for robots than for people.

So, you know, one thing that we’re excited about is we can produce a certificate that shows that literally every bolt on the site was torqued to within the proper specifications, which like, you know, a person just can’t really do very easily. Are you considering making an electric model? Right now, we’re mostly not focused on building construction equipment; we’re focused on using off-the-shelf stuff wherever we can.

I think it would be very cool if we could use off-the-shelf electric construction equipment, but we have no plans to build that in the immediate future. And then I think there’s one last question: I’ve seen some projects that are also trying to automate the module installation with robots.

Are you thinking about something like this in the future? Yes! So, when I alluded to the robotic factory that we’re building towards the end of my presentation, that is the step that will be doing the module installation.

So the robot that I showed you in that video was really a prototype that did one thing well, which was move materials around. Our goal is to do complete mechanical installation of the site, so that includes eventually doing pile driving, installing all the racking structure, installing all the modules. So that’s the vision of the company. Cool! I think that was it.

Thank you guys very much! And again, if you're interested in working on this, feel free to shoot me an email at max@chargerobotics.com, or just go to chargerobotics.com. We have the careers page, but thanks again, and next up we have Oliver from Impossible Mining.

Hello everybody! I'm super excited to be here. I'm Oliver, the CEO and co-founder of Impossible Mining, and we are building underwater robot vehicles which collect battery metals from the seabed, and we're doing it in a way that really doesn't harm the ecosystem that lives down on the seabed. We're actually a B Corp, and so that’s really much very much in our DNA.

So if we talk a little bit about the problem, I mean, we want everyone to drive an EV; you know, we want to move away from fossil fuels and gas. The good news is there's a lot of growth coming, but the downside is that all of this needs a lot more critical minerals—battery metals, in fact. An EV needs something like six times the amount of metal of a traditional gas vehicle, and that's really driven by the battery.

So we've got a need for all of this material whilst we also want to have good ESG—we don't want to hurt the planet in the process of the transition. So the fundamental problem with the supply chain is the fact that we don’t have enough of this material. The ESG characteristics are pretty poor, and the material that does exist in the ground tends to be controlled by China.

A lot of places in Indonesia and Africa are under the influence of China, so here in the West, we really need to solve this problem. So our approach is to really go after what are called polymetallic nodules. I'm actually going to hold one up on the camera here. So this is a rock that forms over millions of years on the seabed, and it's super rich in battery metals.

It actually is the world’s biggest resource of nickel and cobalt. It also has copper and manganese, and there’s estimated to be something like a hundred trillion dollars worth of these battery metals just lying on the seabed on the floor. So we don’t have to blast or cut; we just have to pick them up.

Now, the challenge in picking these up is how do you do it? Our competitors are basically building dredging technology. This is a short clip from their approach. These are massive machines; they get lowered to the seabed floor, and they generate all of these sediment plumes—these clouds of dust effectively that really destroy wildlife, and they indiscriminately remove everything by basically squirting water into the sediment and then sucking it up over this long riser tube. We don’t like that approach.

So what we’re doing is we’re building fully autonomous robots, and here’s a little animation of what we’re building. So these robots will be launched from a traditional shipping container vessel. They're fully autonomous; they have a battery pack, and they have a built-in buoyancy engine that allows them to go up and down. We use a parallel fleet—you can see others returning.

Once they get to the seabed, they don’t actually touch the seabed. They use the buoyancy engine to maintain mutual buoyancy and then they use cameras and AI to selectively pick up the rocks, avoiding any that contain life. This could be sponges or corals or other forms of life, and we actually leave behind a certain percent to maintain the habitat.

Once the payload is full, the battery pack adjusts the buoyancy engine to make the vehicle float to the surface where it’s recovered. Now the payload will be emptied, the battery pack will be swapped, and any maintenance is performed, and now the vehicle can be redeployed on the next mission.

And so, you know, we’re basically building the team in Canada to build these robots, and we’re just in the process of testing our first proof of concept. We were a YC Winter ’22 company, and so we announced just a few weeks ago that we had closed over 10 million dollars in funding.

So we're quite well funded at this seed stage, and now we have to build these proof of concepts. We have four roles that are currently open in Collingwood, so Collingwood is a town north of Toronto, actually on the Georgian Bay. We’ve just rented a very large 10,000 square foot facility, and this is where we will be building these robots.

And so if anybody is in the area and is interested, please spread the word! I would love to have you come and join us and help us deliver this vision of making it easy and less environmentally damaging for everyone to drive an EV.

Okay, thanks! I will attempt to answer any questions. Let me see, let me scroll back. Okay, dimension size of these nodules? Yeah, I'm holding one up. They are about two inches in diameter. They vary a little bit, but they're relatively small. They form over millions of years; there are huge numbers of them.

What's the relative density against the water? They have some weight, so they're just lying, you know, they're rocks. They lie on the seabed, not attached, but they're there. Will they control the food we rent? We have designed our robots so that they will work with traditional shipping containers, and that’s a big part of lowering the cost of collection.

So we need minor modifications to a traditional shipping container vessel, and we also use that vessel to not only launch and recover the vehicles to charge them, but also to transport the payload—the actual modules—to port where they'll be processed to get the metals out.

How deep is the machine able to go? So unfortunately, these nodules, these rocks, they only form in areas of water that have been— that for millions of years, typically three to five million years. So that means they have to be deep. The typical depths are about five thousand to six thousand meters—so, you know, three and a half to four and a bit miles deep—very, very deep.

Are you using RGB cameras or hyperspectral? We’re using RGB. At that depth, there is no light, and so we provide our own illumination. We’re using visible light and potentially infrared and just using traditional images.

How are you treating the biologics on the surface? I’m not 100% sure about this. I mean, there are lots of studies that show there are corals, there are sponges, and there are also other forms of fauna that lay eggs. So our AI will detect these and avoid picking up those nodules. We will also choose to leave a certain percent behind for working with a team of marine scientists to kind of co-design this collection vehicle.

How environmentally economically viable versus others rare earths? Yeah, we think it’s really economic. We want to be about a third of the cost of the dredging, but if you take a brand new mine on land, there are real issues with that.

First of all, in the U.S., it takes eight to twelve years to permit, mainly because no one wants a mine in their backyard, and you often have to displace people—often indigenous people—and so that goes through the courts. So we think the permitting will be a lot quicker.

Secondly, we have a very high-grade resource. This nodule has nickel, cobalt, copper, and manganese at much higher grades than we find on land. All the high-grade metals on land have already been mined. Also, on land, they’re typically very remote, so you have to build a lot of infrastructure. You have to build a train line or a highway to get there. And on land, you don’t typically find four metals in one, so all of these things will come out to be much less costly.

Who owns the rights to the minerals? Great question! There are two jurisdictions. If it’s within the exclusive economic zone of a country, i.e., 200 nautical miles off its coast, it belongs to the country. If it’s outside of that, it’s in areas beyond national jurisdiction or the high seas; it’s regulated by a United Nations body called the International Seabed Authority or the ISA, and they were established in the 1990s.

They have issued something like 32 exploration permits. Are these in international waters? Yeah, okay, so yeah, if it's two jurisdictions, you’re in international waters, it’s regulated by the ISA, or you're in the jurisdiction of a country like the Cook Islands that has established their seabed minerals authority that is issuing permits and regulating.

Can these rocks be treated in existing facilities to extract the minerals, or do you need new facilities? Great question! You need some modifications of the traditional refining, but actually, we also have a team that is trying to do a really innovative way to do this using naturally occurring, breathing metal-breathing bacteria. So we have a team in LA that is attempting to do that.

Why are you not using sonar? We actually do have sonar on the vehicle, and that's mainly for when it is being deployed, to make sure that it doesn't hard impact the seabed. It's for maintaining neutral buoyancy, but specifically we feel that RGB camera works better for the vision system and picking up, and the fact that we control our own lights means that it’s relatively easy.

How long are you out to see? How much material do you harvest per hole? We will do about three million metric tons in a year; that will generate about two billion in revenue. We have a cycle time of about two weeks, so every two weeks the shippers are coming and a new crew is going on station.

Won't you cry extra metallurgical isolation? Yeah, we talked about that. We have something called bioextraction that we're working on to do that. Why Collingwood? Incredible engineering talent. The University of Waterloo is very close by, and that's where our CTO went.

Also, access to the Georgian Bay. Are environmental purposes of these rocks part of an ecosystem? Yes. Life has evolved to use them. As I mentioned, there is life that uses them to lay eggs; it’s the only hard substrate, and so it’s really critical in our approach that we don’t remove all of them. We leave a certain percent behind so that life can exist.

Do you intend to use lbs or other spectral density analysis? We don’t need to choose which rock to pick up. The percentage of the metal in each rock doesn't vary from one rock to the next; they’re all pretty much the same.

So the distinction to pick up the rock or not is based on whether any life is on it, or we want to leave it behind. Good on you for knowing your numbers! Where are you far away from building the prototype? We built for YC at the start of this year. We built the arm in the test tank, and we now have the shallow water proof of concept in the water. We are hoping to be able to demonstrate that in the fourth quarter of this year.

Okay, why seven percent? I don’t know what that refers to, but anyway, I should probably stop there and hand over. Please feel free to reach out to me. My email is oliver@impossiblemining.com.

In this next portion of the event, we will have two quick pitches to hear about two other climate companies. We'll start with Inside Muji Meats.

I am Inza, I’m CEO and co-founder of MujiMeats. As you probably know, meat is one of the biggest drivers of climate change, but fortunately, nowadays we have meat alternatives. You’ve probably all heard of companies such as Impossible and Beyond, but you might also have heard that on top of that, lab-grown meat is a big emerging trend.

The entire meat alternative industry has one big problem, though, which is they can't produce good whole cuts. So if you go to your local supermarket and look at your real meat section, you see like steaks, fillets, chicken breasts, whatever. But if you look at meat alternatives, you will find only ground meat and maybe some sausages and patties.

Muji Meat is a habit spin-off that develops a scalable 3D printing process to produce whole cuts at mass production level, and just to give you a sense of how cool the technology we're working on is, we do have a corresponding Nature publication that completely blew up, and it broke the record for the most number of readers a Nature paper ever had in the first month after publication.

And yeah, we are a growing team; we hire a large variety of backgrounds from electrical engineers to bioprocess engineers to food scientists. So if you have any of these profiles or are just generally interested in our mission, I'm happy to get in touch with you.

Does anyone have one question for Inza? Can you please describe what the printer is like? Okay, yeah. So there's two main differentiators that are relevant for us. So for once, you probably know 3D printers; most of them have only one nozzle per printing hat, and we print with more than 250 nozzles at print per printing hat, which sounds very easy, but there’s a very complicated pressure regulation system behind that.

The other main differentiator is that we can do the material switch within each nozzle up to 50 times per second, which is much faster than if you only have one material per printing hat and always need to relocate that. And it also enables much more precise structures and textures, if that makes sense.

And next up, we have Josh. Hey everybody! Oh yeah, great to be here. Thanks for allowing me to present. I’m Josh Santos, the co-founder and CEO of Noya, and Noya converts existing cooling towers into carbon vacuums. If you have never seen a cooling tower before, they look kind of like this.

For our purposes at least, they’re basically just big boxes with huge fans that move lots of air. We take advantage of this moving stream of air sitting on top of a facility that has already been constructed to capture carbon dioxide directly out of the atmosphere. We do that with a machine that looks kind of like this.

Our machine consists of a few key components that we are developing entirely in-house. We’re developing something we’re calling the contactor, where air plus our carbon capture material contact each other and come together. That material gobbles the CO2 up as it’s passing through our equipment on its way into the cooling tower.

The second thing we develop is a way to move the material from the contactor beyond to where it needs to go. Our material is a solid, so we're basically using some fancy conveyor belts. Think conveyor sushi, but for climate. The last piece of equipment we’re developing is something called a CO2 regenerator. It’s called that because it regenerates the CO2 that we capture into a pure stream. It also regenerates the material that we are using so that we can use it again to gobble up more CO2.

I would be a bad CEO if I told you that this is exactly what it’s going to look like; it’s not—it’s the vision, but we are designing to this, and we just finished our first build, and we're hiring lots of folks.

We got different kinds of engineers like chemical, mechanical, electrical, manufacturing, controls, and we're hiring some materials scientists as well. We're hiring the electrical engineer actively, and the rest are coming later on this year. So if you're interested in cooling towers or carbon capture, let's talk. You can see my email there, and that's probably my 60 seconds.

Does the load change during use? I'm not sure exactly what load you're probably referring to—the cooling load or the electrical load required to cool? And then where are you storing the solid end product? I know I said one question, but… all good!

So the cooling tower with our equipment retrofitted is designed to be able to perform up to 95% or more of its peak cooling capacity. We're able to do that because of the specific type of material that we're using that essentially provides both a low pressure drop across the material and also a high surface area as air is flowing through it.

We actually store CO2 in a gaseous form, excuse me, in liquid form. It's regenerated as a gas, we then compress it into liquid, and we store that in a big tank in the service area of the buildings or industrial sites that we are retrofitting.

[Music]