QuantumScape Corporation (QS) Earnings Call Transcript & Summary

Okay. Good morning, everyone. Welcome to day 2 of UBS Virtual Solid State Battery Day. We're really pleased to have QuantumScape presenting in this session, and we have Kevin Hettrich, the CFO, who's going to take us through some slides. His slides are up right now on screen. So please have a look at your screen to follow along with the slides. The format today will be, we'll go through the prepared remarks, say, 20 to 30 minutes, and then we will open the line for Q&A. As usual, you can feel free to e-mail myself or Sherry with any Q&A that you might want us to ask. [Operator Instructions] And with that, I will hand the call over to Kevin.

Tim, thank you kindly for the invitation to participate today, and good morning to everyone. What you see here is our select set of slides from our investor deck, and you all should feel free to download those from our website. So just to kick things off, QuantumScape was started 15 years ago with the goal to develop significantly better batteries on all the dimensions that are important to consumers and to help commercialize them. We've been focused on the automotive space. And over that 15 years, the first 5 years or so, we're focused on finding material by which we could commercialize a lithium metal chemistry. We'll talk a little bit later about why that's so important and how it links to all the benefits that I mentioned. In 2018, we announced a public relationship with Volkswagen based off the testing of our first small single-layer cells. Over the coming years, in 2020, we started doing full commercial sized parts. We were multilayering it, kind of culminating with our first automotive A samples at the end of 2022 that we shipped. Two years later, we did our first automotive B samples in October of 2024. And then last year, and there's some more detail on the next slide, we did our first vehicle debut, where our first product, the QSE-5 cell powered a Ducati V21L race bike across the stage at the Munich Auto Show, which is a pretty emotional moment for the company, having started out 15 years ago, not knowing if material existed that could meet automotive requirements to then having it be in a first vehicle demonstration. And just last month, and I can say this for a few more days, just last month, we inaugurated our Eagle pilot line here in San Jose, which is a highly automated line that makes those QSE-5 cells. And then here is a shot from that day of that Ducati coming on to stage. What's so nice about a bike as a program for us is that it's a very demanding use case. If you think about the acceleration and deceleration of a bike going around the track, that's a fantastic opportunity to show off the power capabilities, volumetric energy density would translate into more range and then safety, the pack is literally between the legs of the driver. And then being a smaller vehicle that helps us with more end in terms of the accelerating the rate of testing and learning. In this particular project, Audi did the packs for us, Ducati did the bikes. And then together with PowerCo, we supplied the QSE-5 cells. And importantly, cells in that pack, there were cells that had come off of our Cobra separator process, which we had just baselined a month or two earlier, which kind of shows off the speed with which we can go from baselining all the way into sample products. So that was a fantastic moment for the company. And then in the lower left-hand corner, you can see one of the separator sheets, you can see one of the cells, and you can see a battery module of QSE-5 cells. So here's the QSE-5 announced in October of '24. The combination of volumetric energy density, gravimetric energy density and safety in the same cell is unmatched to our knowledge on the earth, 844 watt-hours per liter, over 301 watt-hours per kilogram cell discharged in 12.2 minutes and then has wonderful safety properties. We have mentioned in other context, you can heat the cell up to a few hundred degrees C and you don't see a thermal event from the cell. And I alluded to lithium metal chemistry. The reason why we targeted this and we've worked so hard on the commercialization path to bring it into the automotive market and into other markets is because of this chart here. So on the left, you have a -- let's see, this is the left, you have a single layer cell from a conventional lithium-ion. On the top, you have an anode. On the bottom, you have a cathode. In between, you have a separator, current collector of both sides. When you charge a cell, the lithium goes from the cathode to the anode, it's a growing ball uphill. When you discharge the cell, you want the energy back. It's like the lithium goes from the anode back to the cathode. It's a growing uphill and you get it back up and back. The benefit of a lithium metal cell is shown on the right in the version we do, which is the desirable variant, as manufactured, there is no lithium anode, we call it anode-free as manufactured. The current collector touches the separator. On first charge, lithium that's already in the cathode material that you buy, just like anything else, plates out on that first charge. And the elimination of the volume and the weight from the anode, if you compare the column on the right with the column on the left, you can imagine the volume and weight savings, that's volumetric and gravimetric energy density. If you imagine you lithium-ion going from one side of the device to the other, the distance you're traveling on average from the middle of the cathode to just the other side of the separator, you've roughly reduced in half. And so that's the charge benefit that we talked about that enables that 12.2-minute charge of QSE-5. In terms of life, one of the major sources of life loss occurs on the anode of the conventional cell where the liquid interacts with that graphite silicon structure and forms a layer of what's called a solid electrolyte interface. As that cracks and reforms over time, you're chewing up active material and adding to resistance to the cell. We don't have an ion conductor in the anode. It plays on the other side, nor do we have graphite silicon type structure. So the elimination of that major source of life loss, we credit with the wonderful capacity retention that investors have seen in our A-sample testing. Safety. So we talked about energy density, we've talked about charge time. We've talked about life. The other 2 elements will be safety and cost. Safety. The separator itself is an organic material. The anode is flooded. It is an organic material and is flooded with an electrolyte, which is also an organic material. So we replaced the conventional separator with an inorganic ceramic, and we eliminate the organics from the anode side. So it's that reduction of organic material and then the inclusion of all these inorganic separator sheets throughout the battery that we credit with the safety performance data that we've shared with investors. In the QSE-5, there is one organic material in the cell that's in the cathode that is the electrolyte we use for ion conduction. So we've substantially reduced it. In the QSE-5, it's not -- there is that one lone material, but the reduction we credit with the safety improvement. And then finally is cost. At scale and at maturity, we believe while delivering that performance, the cells will also be lower cost because the savings from the anode, both the graphite silicon you're eliminating combined with the liquid electrolyte is more savings than would be required to replace the separator with our -- the conventional separator with our ceramic part. And that's the 5 major benefits that people care about and hopefully, a little bit of intuition with how they link to our architecture. One final comment is that our choice of cathode material would be something we would tune and to the application or to the preference of our customer partners. Across the board, with the initial prototype we've done, the vast majority of our test data is on a nickel-rich NMC because our customers want to push the limits of what's possible in their products. We have shown historically data as well on LFP, which is a different product positioning, also intellectually interesting where you're pairing the world's cheapest anode, which is nothing with the world's cheapest cathode material, but more on that to be discussed in the future. Dan had just done a bit of a peek reveal of how is the QSE-5 compared to multiple automotive cells on the road. What we've plotted here on the Y-axis is charge time, which is shrinking as you go up as well as volumetric energy density on the x-axis. And that forms kind of a power to energy frontier. And the QSE-5, our goal in the QSE-5 and subsequent generations is to keep pushing that frontier further and further out. In any cell chemistry, you can trade off those 2 dimensions generally by making the cathode thicker. You improve energy density and you hurt power and you can kind of go up and down that line. Our goal with the QSE-5, we pushed that line out. And with subsequent product development, the green would be a larger form factor using the same technology of the QSE-5 is to keep pushing that line further and further out. You actually don't see LFP cells on this because we started X-axis at 500 watt-hours per liter, they would be plotted off of the chart because that's not a very energy dense cathode choice. I mentioned -- I alluded to you get these benefits by the elimination of the anode line and a conventional factory does 4 things. It makes anode, it makes cathode, it assembles the cell and then it test them. So I'll talk about 2 of those areas. So we don't need to make the anode and electrode. That's -- of course, that's the source of many of our benefits and cost savings. The other thing to highlight is that we don't need to form that solid electrolyte interface layer either. So in the formation and aging step in a conventional lithium-ion plant, the equipment that does that, that forms that solid electrolyte interface layer often occupies 15% to 25% of the floor space. And that's a capital-intensive step. It's time intensive and you're tying up a bunch of inventory. So it's actually also one of the most expensive steps in the factory, and we substantially reduced the time required because, as I mentioned, we don't need to form that solid electrolyte interface layer. So those are the 2 major areas where we expect cost savings. And it's not enough to have performance in the cell itself, you need to hit the scale and the quality and the cost points of the automotive industry. Our core IP and our unique part is that ceramic separator. And the 2 elements that are required in order to do so at scale and then at maturity to, in fact, outperform lithium ion at cost. is to make that separator very cost effectively and scale and quality. So what -- in ceramics manufacturing, the most expensive step is often the heat treatment step, and that's the area where we worked to innovate. In 2023, we were on a continuous process to make our separator that tends to be where you'd want to start with cost-effective processes that they'd be continuous. And working with a vendor to modify their tool, we sped it up by almost an order of magnitude, and we call that our Raptor process. And then last year, approximately in the summer, we base lined our Cobra process, which is yet another order of magnitude on top. If you combine the improvements of Raptor and Cobra together, they're roughly 200x faster in terms of heat treatment time than the continuous process from which we started. So that type of innovation is transformative for the cost of making our separator part because you're driving so much throughput out of the most expensive step in ceramics processing. So that's the process part. And then QuantumScape will talk about in a minute, we've adopted a licensing model where it's our job to innovate, to take something to a pilot level, B-sample level of maturity and then off of that line to win customers, to strike licenses and to also enable the rest of the supply chain. The reason I bring that up is last year, we also announced partnerships with arguably the 2 -- the world's 2 leading ceramics players, being Murata in Japan and Corning here in the United States. And we're very excited about that partnership and that we think we both bring a lot to that ourselves with the combination of materials and equipment and process that makes these separators, which enables the performance that we talked about. And then Murata and Corning, you have world-class manufacturers who use the example of the multilayer capacitor industry, the MLCC industry, they make tens of billions of MLCC capacitors and those costs a fraction of a penny or a small number of pennies a piece. And just they are amazing products with their thickness control and uniformity. So we couldn't imagine better partners to work with industrializing and taking this to scale. And then we included -- this slide is a favorite from investors, one in which we get a lot of questions. We think the data is the best place to start. And what we've done is we -- against this chemistry, lithium metal chemistry, which has the benefits that we discussed in terms of energy density and charge time and safety and life and at maturity and scale cost as well. We've plotted the prototype data around the world that we're aware of, on the Y-axis is charge time. So this is the charge time you use in repeat cycling. On the X-axis is the cycle life until 80%. The size of the circle is correlated with energy density. If you have no lithium -- if your anode-free is manufactured, you have a big circle as an approximation for energy density. If excess lithium is applied, we give a smaller circle because that would hurt the energy density. And then finally, the color. Red would be something that has temperatures and pressures that exceed what's used in the automotive application. Blue is you're kind of in the ZIP code in the automotive application and what an automotive system could provide. And then green means you can operate at room temperature and no applied pressure and then you have the full range of applications open to you, including consumer electronics and -- any other type of application, there's not a system around supporting you. So you see based on the data, we are far ahead on multiple dimensions simultaneously. And for us, looking at this chart, you actually have a combination of different materials being used. You have their ceramics and polymers and sulfides and each has their own independent challenges to them. And outside of QuantumScape and the ceramic we use, there is no proof point of -- to date of another material matching the types of performance that we've demonstrated. And then just very briefly before we wrap it up, we've been working with Volkswagen since 2012. And that's the Volkswagen Group, which has many of the world's most iconic brands, we kind of show on the lower left corner, like Porsche and Lamborghini and Audi and BUGATTI and Ducati, et cetera. They've been an investor in multiple rounds. We're a JV and then later licensing partner, the Head of the Center of Excellence serves on our Board as 1 of 2 VW members. In 2024, we struck a licensing deal with them that goes up to 80 gigawatt hours and put a team from PowerCo here on site to develop and bring up the Eagle Line together. And then last summer, that deal was expanded and enhanced to go up to 85 gigawatt hours, that last 5 being usable by PowerCo outside of the automotive sector and then giving the company up to $130 million of payments we can earn by achieving blocks of work that are laid out in the agreement. The world saw the first of those, which was that debut at the Munich Auto Show. In this VW, there's a nice kind of pipeline of many of the who's who of global OEMs. The way we think about it is there's kind of a technology evaluation agreement where you're having engineering discussions. There's a more mature version where you're actually doing joint development and joint research together. And then the final node and maturity is what we've created with Volkswagen, which is that full-blown collaboration and licensing agreement with the magnitude of payments that we have. And we'll talk about the goals later is to continue to execute on the Volkswagen relationship while we advance those other automotive relationships. And the only other final thing I had is, and as Siva mentioned in the recent blueprint video, we've received a lot of inbound interest from nonautomotive applications and particularly applications that put a very high premium on battery performance. Some examples we've talked about include aerospace, defense, robotics, AI infrastructure, et cetera. And one of the 4 public goals that we talked about was to go beyond automotive into these spaces. But Dan, if you go back just very briefly on that licensing example. So we're a licensing company, as I alluded to. There are 2 major sources of cash flows from customers that are possible. The first under the collaboration phase is we would take a technology platform like the QSE-5, here's a QSE-5, here. And we work with the customers to say, what's exactly the dimensions you want? What's the exterior form factor that you want? Do you want to play with the choice cap thickness, et cetera? And we would do development that is tailored to exactly what they want. And in exchange for doing custom development for them and the up time in our pilot line and our own internal resources, we would seek a cash compensation for that. What's nice for investors in that model is that this is a far more asset-light model than doing the manufacturing yourself. We have to sink very high sums of capital into the ground and then bring up the factory before you actually return cash flow to investors. In this case, we're actually able to receive from customers cash flow quite early in the process, which we see as in-period validation of what we're working on is important, and it's an important source of cash flow, especially in a success case as we line up multiple automotive customers at the same time with active collaboration payments. The incoming cash can get exciting. Longer term, by and far, the larger economic opportunity is when a customer takes the license and is producing out of their factory cells incorporating the technology in the license. So at that point, it would be a high-touch licensing model, we support our partners however we tend to make that transfer successful. And then that would be where the bulk of the economic opportunity would come from. And then as a technology development company, it's our job to continuing to push generation after generation on a rhythm and to develop the materials, the process and the equipment as well as the supply chain. And you saw the first 2 examples of publicly announced partners in the supply chain in Corning and Murata. And in the future, we look forward to sharing additions beyond that. I think that's -- I think we wrap it up with what we're focused on in 2026. We've had a great track record of saying what we're going to do and doing what we said in terms of our public goals. So we look to keep that up in 2026, where our goal is to do 4 things. One is to demonstrate scalable production on that Eagle Line that we just brought up. The second is to advance automotive commercialization with Volkswagen. Siva mentioned field testing with that V21L race bike this year and to continue to progress those other automotive relationships forward. I touched on briefly the new high-value markets. And as I mentioned, the QSE-5 is the first step in the road map, and we will talk more about what's coming next. So that's QuantumScape at an overview level, and I'd be happy to shift to questions with you at Tim as well as at the end of the call, any questions that the broader set of investors have.

Thanks, Kevin. [Operator Instructions] I've already had some questions come in over the line. So I will kind of kick off. Yes, the chart that you showed with the green and the red, basically the dot chart that you showed was -- did attract a lot of attention, right? And we kind of had a -- like yesterday, we had a lot of talk about pressure and that being a problem because if you -- it can add weight and cost and be a drag on energy density. So when we think -- and it seems like the other players we spoke with were using pressure to inhibit dendrite formation. So is -- in the case of QuantumScape, is it the ceramic separator, the key -- I guess, the key component that's inhibiting dendrite formation and allowing us to avoid this kind of having to have the cells operate under pressure?

That is a very good question. So first, I preface it that your conclusion is right and that we think for automotive, maybe a few atmospheres of pressure. We define it here as 2 to 5, you could kind of think of a pack design around it. What's more typical, I believe, is closer to 1, and that's where you see we've taken all of our testing for automotive. And then we've shown no applied pressure as well. And that isn't in our core baseline, but we've shown our capability to do that and something we'd like to bring into our baseline. So the advantages being the ones you mentioned, the cost, the complexity, imagine just like the challenge of pack assembly at very high rates of pressure. And then if you think about, yes, you may -- in a system where you use pressure, that might help with dendrite resistance or suppression, but just imagine what that's doing to your components over time. We think that can also hinder life loss. So we set as a goal to effectively do the same thing as what our customers are used to lithium-ion and we achieve that. And how do you -- it goes to the separator and it goes to the interfaces with the separator on both sides, and it goes to the cell construction. So I'm going to hold up a QSE-5, and I'll see how we can closely get here a little bit of an angle, that doesn't focus. You see there's just an imperceptible like little ridge around it. And what that ridge is doing is that the difference between the charge and the discharge state might be about a millimeter in this design. So the frame doesn't move and that little bit of expansion is actually accommodated inside that frame. So we call it the flex frame. We can also do this in a prismatic cell, imagine a hard prismatic cell, the frames inside and the exterior dimensions of the prismatic cell don't change. So that's how we handle the expansion. We don't change the exterior dimension. So we're not trying to fight the cell. It's actually expanding within the volume that we've kind of created for it. So the combination of that, the separator material and interfaces on both sides, it's really all of those things together, and it was not a small amount of work. And what we found is that over time, as an idea -- if it's an idea, a cell design, a process, a step as it gets more mature, it actually gets simpler and more elegant. And this is an example that we think is actually a very elegant solution to controlling volume expansion.

Thanks, Kevin. Helpful explanation. And going to the manufacturing schematic that you showed earlier, so that we don't have -- the anode side of the process is taken out. But on the cathode side, and you also spoke to the formation with the savings that we would have with eliminating expensive or long formation process. Are we able to use in terms of like the mixing and the calendering and the slitting and so on, is this kind of able to use existing lithium-ion infrastructure? Or is a lot of this new equipment that's specific to your process?

At the zeroth order, the answer is yes. Like the cathode, the materials, the process is very similar and uses the same or very similar equipment throughout. And so the answer there is yes.

And on the anode side, like when we talk about anode less, are we still -- are we -- in the in-situ lithium metal formulation that we're doing in situ, is that forming onto a substrate like a copper foil current collector. So we still have some type of current collector that...

Correct. So it does form on to a current collector, and then I'll get a little technical on this one. So what's very interesting about conventional lithium ion, same thing is on first charge and that structure on the column, I need to do the reverse here on the -- hopefully, I'm lining up the same way for investors. In the conventional cell, when you charge at the first time, the lithium goes into the structure for first time. Same thing for us. And in order to accommodate for the SEI layer, that solid electrolyte interface layer, they put more lithium in the cathode than you actually need to cycle back and forth because they're planning for a moderate amount to get lost on that first cycle. So what's neat about us is that on the first charge, we do that first plating and never again is that current collector bare because that first charge has a bunch of excess lithium in it. So that's -- it's a pretty -- it's a beautiful fit with the existing cathode material that's effectively over-lithiated to accommodate for that SEI formation. So we do on the first charge in our own factories, that first formation step and then -- which is, again, a tiny fraction of the time of what it takes to do an SEI, but we would do that and then you would never see a bare current collector again in normal operation.

And like recently in the presentation, we've been talking to some of like lithium metal film type players. And we kind of -- in terms of the upstream scaling of lithium metal, there seem to be real issues. But for this analyst design, are we just using commercially available either hydroxide or carbonate, so we don't have to worry about sourcing upstream lithium metal?

The simplest was buying the cathode material from the cathode manufacturers, whatever precursors they think are most cost effective, make the supply chain most robust, lead to the highest performance. It's buy the conventional cathode material basically is the answer to that. And then the question for the advantages are very clear. It not need to start with the lithium metal foil, cost, energy density, handling during manufacturing. It's a big advantage and would be a very big headwind to overcome in your cell design and competitiveness if you wanted to include it. You really shouldn't need it because there's more lithium metal than you need to reversibly cycle in the cathode material that you buy.

Got it. [indiscernible] I see you have your hand up. [indiscernible] please unmute yourself and the floor is yours.

This is [indiscernible] covering the automobile space. So Tim discussed about the pressure, I want to discuss the same topic on the mass production side, okay?

Yes.

So can you please explain what's the advantage of the uniqueness of our Cobra process? And especially I'm interesting in the pressing process like pressing the machine, the partner with machinery company also [indiscernible]

So there's a lot of -- I -- so let me -- I'll show what I can. There's a lot of IP there. So we use the names of 2 fast animals, a Raptor and Cobra because they're speed. So the difference that we care about is the heat treatment time, that centering step is how quick can we make that -- so a given tool doing it is as productive as possible. And that's to think about depreciation, think about square footage used in the factory floor, think about the energy, any type of consumables, et cetera. So our desire was to make that as fast as possible so we could drive throughput and take down all of those other costs associated with heat treatment step, which is classically substantially the largest cost step in any type of [indiscernible] manufacturing. So how we do that was, I think, quite innovative. And to be candid, that was one of the things in addition to the business opportunity that attracted Murata and Corning is they're also really excited at that process that was innovative to them as well, and we're excited to see how much further we can take that together. So it's related to speed. And the Cobra, we think of broadly as the process to make the separator, but the real innovation is in the speed of heat treatment there. The only other -- final thing I would say, as we've shared before, is Raptor was done with the modification of an existing continuous tool or work for closer to the supplier. Cobra having shown that, that's a vector we can do, we said how hard can we crank that vector, we did some proofs of concept in parallel with the Raptor tool to show that it could be done. And then we worked with a supplier to actually do a tool that's more special that actually implements those proof of concepts into a pilot line level tool, and then that's Cobra. So there was roughly a year between generations there, and we baselined that successfully sometime around June of last year and cells made off of the Cobra process went into that bike that went across the IAA stage.

So what do you think is the current stage of like to reach the ideal mass production -- a step as mass production process because there are lots of debate of testing process like WIP, HIT or ISP, there are still, I think, is not the standard for how to produce the solid state batter. So what's the current phase of QuantumScape?

So in terms of maturity, I would say you're at -- we've been producing B samples. We're on an automated -- a heavily automated pilot line here in San Jose. And where we are is that was just announced last month. So it's been kind of brought up and all the things kind of worked out towards the goal of the demonstration. That line does 3 things. One is to produce a whole lot more parts that's useful for internal development. That's kind of the first goal to continue to mature, to root cause, the types of things you work out to drive uptime and yield and reliability, all those things that you do with systematic, methodical, iterative process development. That's goal one. Goal 2 is to make a whole bunch more parts for Volkswagen, for the other automotive OEMs in the pipeline and beyond automotive. And then the third is this is the basis of tech transfer. So following success here to use the case of Volkswagen and PowerCo, they would then order a larger equipment set based on what they would learn here, and we use that as an opportunity between generations to make improvements. So Thomas Schmall during the Munich auto event, Board member, in charge of all kind of technology and sourcing, said that the goal together was to get to large-scale commercial production before the end of the decade. So he didn't put a pin on exactly what that meant. But regardless of how you define it, there are still years of work to do together and execution to make that happen. And we just happen to have been on a good pace of good execution. But there's very much work to do together with our cell manufacturing partners like BW PowerCo, our suppliers of components like Murata and Corning, who we're working with and then all of the material and the equipment manufacturers we're working with for kind of generational improvement. So I don't want to underscore that there's still a lot of work to do. It's just that we're at a pretty exciting kind of moment and have been making good progress. If you have some question, go ahead.

I think Sherry, could you please unmute yourself and the floor is yours.

Thank you Kevin. Thank you, Tim. I actually got a question from client, and he's basically asking what is the progress of partnership with other OEMs than Volkswagen Group?

Excellent. So we did a press release in December. So if you think about just 2025 is a very good year. We upgraded the Volkswagen licensing and collaboration agreement. And we took 2 major OEMs to that precursor stage of what we refer to as a joint development agreement or a joint research agreement. So the intent behind both is following success, if we can build confidence with the execution of that joint development and joint research scope, we would then have mutual interest to move on to that next step, which would be to try to sign a mutually advantageous collaboration and licensing scope in the same magnitude of Volkswagen. So you think of we have a relationship with a number of OEMs. There's geographic diversity, there's size diversity. You have pure play, you have kind of combination of EV and combustion engine. And you have Volkswagen Group by far is the most mature between the collaboration licensing, the Board members. We've -- it is important to note that the last 2 quarters, we've collected just under $20 million of collaboration payments. roughly, we've had an average of about $10 million a quarter. We do note every time that those payments and amounts can be lumpy quarter-to-quarter following the work. But we are in the -- we are kind of taking IP and doing custom work for VW and kind of getting paid for it. So that's a pretty good view in the funnel. There's a number of OEMs in that kind of agreement stage. Then we have 2 in that joint development agreement stage and then we have Volkswagen. And our goal in 2026, the second one I mentioned is to continue to advance those partnerships.

And the second question would be, do you happen to have any cooperation with Gotion, which is a close partner with PowerCo and is based in China?

Nothing publicly announced. We do not have a public partnership with Gotion.

Thanks, Kevin. We've had a -- I think the slide that you have up now about agreements in other sectors is really interesting, and it's been pretty topical during the conference...

Yes.

I guess, one, we've heard a lot about the National Defense Authorization Act and the opportunities that, that might create for unmanned drone or dual-use tech. And then -- okay, maybe we'll just leave it at that question first, and then I have a follow-up.

And then your -- the question specifically around kind of defense...

I'm sorry -- the opportunity for -- is the opportunity in consumer, maybe not just defense, but in general, is the opportunity that...

Yes.

You're looking at in consumer like...

Absolutely. So I would take the agreements where we have them and then the more recent inbound interest would be from defense, aerospace, robotics and AI infrastructure. So if you put all those together, I think that's some of the most exciting nonautomotive opportunity landscape. And then in defense, I think there's resonance with 3 things. You have the energy density advantage, which translates into payload or flight time. You have the charge time, which can keep your drone using it more often in the air than on the ground. And finally, safety. If you're in a defense application, the majority of drones that are transferred by navy ship or the safety requirement during that transport is incredibly high. As you can imagine, you don't want a fire breaking out on a navy ship. So the QSE-5s combination on those dimensions has been getting inbound interest. If you wanted to watch Siva's recent advanced battery blueprint video, there's actually some nonautomotive application footage, including some drone footage in there.

Cool. On robotics, we had questions about this particularly for untethered robotics. Why is lithium ion not good enough? And what -- is it about weight or volume or operating time, so like maybe power or cycle life? Or what is -- where do you see the -- yes.

Yes, great question. I was told this week at the International Battery Center, IBS, the -- so multiple things. So the cost of batteries is about 1/3 of the cost of the humanoid type robot. So cost is certainly important. Volumetric energy density, you get more useful time out of the humanoid. Hopefully, this is a house-based humanoid that's helping to clean and cook. That would be a wonderful thing. I hope that world comes quickly. So volumetric energy density would be actually usefulness, charge time again to keep that robot working as you want. And then imagine in a home, of course, the safety feature is another before getting to cost. One question we've got is like, why the people question like the automotive focus, like, hey, it's kind of cyclical back and forth. I would just reiterate that we remain very automotive focused. It's a massive market. I don't think it's going anywhere. I think autonomy in cars actually is going to accelerate the EV penetration. But most importantly, if we're successful in automotive, it gives you the scale where you can bring that cost point to the other application areas. So that would be that fourth element where can you bring down the cost of solid-state battery that you then put into a humanoid robot. And those humanoid robots for any husbands out there, I'm a little worried that there's a little bit too helpful. And then if they have that ability on my Tesla, you can kind of pop on that -- you can pop on Grok and there's different settings, therapist settings. So just imagining this very helpful humanoid robot going around who also has like excellent listening skills. I do see that as a little worrying for all of us, husbands are going to up our game.

And on the storage side, I think we started our conversation in the prep talking about the Korean, all these battery capacities in the U.S. that are being shifted over to LFP for storage. Is the storage opportunity for QuantumScape like on-site, maybe the BBU where safety and maybe sea rate is really important. Is that where the interest is?

There's increasing interest to be close to or in the rack where you can both combine backup combined with an ability to handle peak power. And you can imagine the benefits there is the real estate is incredibly expensive. So you want volumetric energy density use as little as possible. Power, you don't want to derate that very expensive hardware that you have and safety kind of goes without saying that you want that to be safe. So for that emerging applications, which is being enabled by these AI racks are maybe consuming 20x more power than their predecessors. So this idea of how do you think about getting power to them cost effectively and efficiently and the rules for different energy storage devices, be it lithium-ion battery or capacitor or combinations of them, that is an area where we've also gotten inbound interest.

Interesting. It would sound like that we would kind of -- with this idea that we would use, we would have the backup and then some ability to handle peak power on site would lead to more duration of batteries or larger batteries on site than what we might have thought.

The other thing from an infrastructure operator is you don't have to build out all your infrastructure at the same time. You don't have to size it for the end facility size for kind of peak power, you could actually build it up gradually as you install infrastructure, which is also nice.

Thanks, Kevin. And then from the investors, I guess, final chance here. If you have a question, please use the raise hand function, we're going to try to wrap up on time like within 3 to 5 minutes. So if there's any question with investors, please use the raise hand function. Give that a second. Okay. Well, it looks like we don't have any further questions from investors, and we'll give back 2 or 3 minutes. Kevin, thank you for taking time this evening or accommodating the Asia time zone to make this presentation. It was really great. We really appreciate your time. And hopefully, we could hear from you again when there's -- when you have another update on the milestone or a major announcement.

I would welcome that. So thank you again, Tim, for the invite. It's a pleasure to be here, and good morning to everyone, and I wish you a great rest of the day, and we hope to report back with further progress when we talk next.

Thanks, Kevin. Have a great evening.

Thank you, Tim. Thank you, everybody.