Kodiak AI, Inc. Common Stock Evercore's 9th Annual ADAS, AV & AI Forum
Review the key takeaways and the transcript of this earnings call.
Transcript
Preview the first fifteen paragraphs, organized by speaker.
Pretty self-explanatory, but— Yeah, yeah.
Have we moved the slides? Oops, sorry. Mic is on. Don, you want to do a mic check? Here you go. testing? Yep. Here you go. Sounds good.
Great. All right, let me know when we're ready to go, guys.
Good? All right. Hello, everyone. I'm John Sager, Director for Global Automobility here at Evercourt ISI. With me today is Kodiak's Founder and CEO, Don Burnette. Don? Awesome. Well, thank you—it's great to be back.
And excited to give you guys the updates and talk a little bit about what Kodiak's been up to over the last year or so. Very, very exciting times in the self-driving space, particularly for us. So for those of you who don't know, we're a self-driving company working on the software aspect, AI-powered, intelligent driving. We really want to focus on the world's toughest jobs. That's kind of our guiding light here at Kodiak. We often talk about the dull, the dirty, and the dangerous. If it's dull, dirty, and dangerous, people either don't want to be doing those jobs or they shouldn't be doing those jobs. And that presents a massive opportunity to inject autonomous technology. We focus on three primary verticals within transportation: we have long-haul trucking—talk a lot more about that later—we have industrial trucking, so this is think about everything that's off-road. Where we've now deployed over 35 fully driverless trucks as of the end of—or 35 driverless trucks as the end of last quarter. These trucks are out there operating with nobody in the cab, 24/7, for our launch customer. This is a DAS business model, meaning the customer owns and operates those trucks.
And we support that operation. We carry multiple types of loads, all the way up to triple trailers. Which is very, very challenging. Even for human drivers. And that allows our customer to increase their ROI in that area. And then, of course, we have defense. We believe that Kodiak has the most commercially mature technology that can be applied to off-road and defense applications as well. Show a little video here. Oh, nope. Can we play it? Maybe not. Can we play the video?
Will it play? Okay. Well, we'll skip the video. So most recently, Kodiak's been focusing on our driver-out launch—our commercial driver-out launch for highway operation. We set up shop in Texas back in 2019, and we've been running commercial loads there ever since. We work with multiple carrier partners across all of the southern United States. And most recently, we've been really focusing on our launch lane, which is going to be Dallas to Houston, which he announced a few days ago. And we're now seeing consistent intervention-free deliveries from Dallas to Houston. So we have some videos. This is a long one. We put up some videos. Time-lapse online where you can watch the Kodiak driver system from depot to depot and to end deliver freight commercially with no interventions, no interactions from the safety driver at all. And we're doing this on a regular daily basis now from Dallas to Houston. We just announced this morning IKEA has been named our driverless launch partner. We've been working with IKEA for well over 4 years. We've driven over 750,000 miles together. That's not the total that Kodiak has driven. That's just what we've driven with IKEA over those 4 years.
And we've delivered over 1,300 loads. And that includes two retail destinations where we actually deliver to an IKEA store. So really excited to partner with IKEA and finally be able to announce that they will be our driverless launch partner at the end of the year. On that note, we've been tracking our progress through what we call our autonomy readiness measure, which is effectively a measure of all the claims in our safety case that we need to complete before the system is validated to a safety level where we can safely pull the driver. As of the end of August, we're at 93%. And we'll be giving monthly updates between now and December as we reach 100, close the safety case, and initiate those driverless commercial runs. We have—it's hard to see on the screen—but we have a network of over 26,000 miles that we operate on. The ones you see in yellow are the ones we do on a daily basis. So we're operating freight daily out of Dallas. Dallas to Atlanta, Dallas to Oklahoma City, Dallas to Houston. Dallas to Laredo and Dallas to El Paso. So our testing and operations facility is based out of Dallas.
Which is the most convenient location for us. But we do have an extensive network going coast to coast, including up to the north of Mid-Michigan, where we do a lot of winter and snow testing. But our primary operations are still in the Sun Belt, which is where we will launch at the end of the year and continue to expand from there. We are accelerating a deployment through our partnership with Rausch. We've selected Rausch as our upfit partner. That they integrate all the AV technology that we need onto a truck and provide that for our customers. We've already set up a bespoke line with them, which we've been operating for well over a year now. I think last year at this conference we were just announcing Rausch. Now Rausch is in full speed. We're brought on a new truck platform that the Daimler Westin Star platform, in addition to our previous PACAR platform. Again, really stressing the modularity of our hardware kit. And system that will work on multiple makes and models of vehicle. And we're ready to accelerate that deployment into 2027 as we achieve driverless on highways and begin to expand commercial operations. I think with that, that's what I wanted to share.
All right. Great. Thanks, Don. So what safety cases are left to be solved for the arm that closing that final 7%?
Yeah. It's a complicated set of claims. You know, we've talked about it in the past. There's all these different aspects of safety: safety culture, safety management system. There's functional safety. There's safety of the performance of the system. People tend to focus on safety of the performance of the system, meaning does it execute lane changes and does it hit the brakes and does it avoid collisions. There's a lot more to a safety case than just the performance. That is one big aspect of it. But really, it's a lot of the validation aspects when it comes to testing in close-course conditions. So there's three ways that we build evidence to support our safety case. There's on-road operation. Obviously, you go out, you drive your vehicle, and you see what happens. That also helps with exposure. So part of the safety case is knowing how often things happen. But in terms of performance, you have on-road development. Then you have simulation. Simulation picks up the bulk of what we do at Kodiak. But we believe that simulation can only take you so far if you truly want to validate your system in the real world. You have to test it in the real world.
And so the third avenue is to go to the track and to test under closed conditions. Here, you can set up a lot of diagnostic tests. You can set up a lot of challenging tests that you otherwise are unlikely to encounter in the real world. And you can also set up tests from simulation to verify that simulation is giving you good results. Now, for us, the primary validation stages that we're finishing up are on the close-course tests. Those are the ones that take a lot of time. And unfortunately, AI still doesn't help you with that today. You know, there's a lot of manual work. You have to send a team out to the track. You have to find tracks that allow you to go to highway speeds with a semi-truck, which is there are not many of them in the United States, right? So you have to make sure you have time. We have now three different locations that we're testing on a regular basis in order to kind of close out all those cases. And then a lot of it comes down to the functional safety testing, validating minimum risk maneuver conditions, fallback safety controllers, backup safety controllers.
Basically, ensure that no matter what goes wrong, that there's always a healthy system that can be there to take over and control the vehicle in a safe manner, whether that's to continue driving, whether that's to pull the truck over, whether that's to call for help. We want to test each and every scenario to make sure that we've checked all the boxes. And that's primarily what we're doing over the next couple of months.
Yeah. Makes sense. So it feels like largely the technical and the regulatory issues have kind of been solved, right? So at our AV trucking tour earlier this year, one of the things that we were kind of surprised by is that even though this makes a lot of sense from a total cost of ownership perspective, it came pretty clear to us that the OEM is kind of the industry's bottleneck. And to your credit, you've got 35 customers and I think more than 40,000 hours of paid driverless operations. So what are the key milestones that demonstrate that you're successfully moving from that early commercialization phase towards scale?
I think that our industrial deployment with Atlas Energy is a really good case study to understand what is the evolution of a commercial deployment. When we talk about this sort of in a vacuum, again, most of the industry for the last two decades has focused on the AI, the software, the performance. But I often talk about the three pillars of autonomy. It's become very clear that the technology is only one of three very important concepts that you need to deploy a successful operation. Technology being the one. Safety being two. And safety should not be overlooked. We already talked a little bit about that. But the third pillar is the product itself. And if I were to hand someone in this room a driverless truck today, and I said, this truck is yours, go forth and make money, you'd probably ask, well, what do I do with it? Right? How do I interact with it? How do I turn it on? How do I get support? How do I tell it what to do? What's the interface? Do you interface with my software system that I have today? Do I have to have a new software system?
FULL TRANSCRIPT
Continue the full translated transcript in StockNow.
Access every statement, the English original, and speaker-by-speaker history with StockNow Pro.
View the full transcript with ProThis earnings call has ended.
Keep exploring
