00:00Robotics, AI and automation are rapidly changing how machines interact with the physical world.
00:06As companies push toward more advanced autonomy and physical AI, the race is on to develop
00:12sensor systems that can help robots and machines see, understand and navigate their environments
00:18more accurately than ever before.
00:20Joining us now to break it all down is Angus Pakala, Co-Founder and CEO of Ouster, Inc.
00:25Thank you so much for being here, Angus.
00:27Thanks for having me.
00:30Of course, so we definitely want to dive in here and learn all about Ouster.
00:33So first of all, for people who are not familiar with the company, tell us a little bit more about
00:37it.
00:38So Ouster is a leading provider of sensing and perception solutions for physical AI.
00:44So we've become a category of one company selling a combination of LiDAR sensors, cameras, AI compute,
00:53and the sensing and perception algorithms that are required to give vision and perception
00:58to the next generation of machines and making them safe and flexible for real world applications.
01:05Yeah, it's just incredible to think about all that goes into these machines and all of this AI,
01:10so much more beyond any part of AI that a lot of us use in daily life.
01:14It's just a chat GPT or something like that.
01:16All of these intense sensors and all this technology.
01:20So tell us a little bit more about the current state of robotics and automation market and how all of
01:25that is right now,
01:26and then how you're changing that.
01:28Yeah, so everyone is familiar with the concept of AI because now we're chatting with AI daily,
01:36but there's another wave of applying AI to real world physical objects.
01:42The machines that move around our world, the cars, the trucks, the combined harvesters in our fields,
01:48the mining machines, construction equipment, drones, you name it, if a machine needs to understand its environment,
01:56the best combination of technology to do that is the new wave of AI combined with advanced sensor technology
02:04right there deployed on the machine.
02:06Because machines need eyesight, superhuman eyesight to both do productive work in the world,
02:11but also do it safely and better than a human, more safely than a human.
02:16So the overall promise here is that for the very first time, we have a nexus of different technologies
02:23that make this future possible.
02:26We have the AI compute from NVIDIA pushing the frontier of edge, edge inference.
02:33We have advanced sensor technologies from ouster, lighter sensors, cameras, and we have the algorithms
02:40that make sense of all of this that we can deploy onto this incredible diversity of use cases.
02:47I think Jensen Wong put it extremely well. He said, anything that moves will become robotic.
02:53It's such an interesting proposition to think about that all of these machines are becoming more autonomous.
02:59And like you're saying with that quote, really just becoming their own category that is something
03:06different than anything we've seen so far. Tell us a little bit more about the LIDAR technology,
03:11because that might be something that people are not as familiar with, but it's such an important part
03:15about these machines being able to be more autonomous.
03:19Yeah. So again, this is all about giving machines eyesight and humans use basically a form of camera
03:27technology as a direct corollary to eyes, but there's actually a better technology that you can
03:34give for machines that gives them 3D vision and that's called LIDAR. So ouster is a leading provider
03:38of this technology. It stands for light detection and ranging. It's very similar to echolocation or sonar
03:46or radar. But basically it's measuring the 3D environment using laser light to give superhuman
03:53vision to machines. And actually now with just the last week, ouster released the very first
03:59native color LIDAR that gives the LIDAR sensors, both 3D vision and color vision all in one. So we're
04:06really giving the full context of the scene through our technology LIDAR with native color that gives both
04:133D information and color information to these machines. And this all comes back to safety.
04:19The goal of perception and advanced sensor technologies like what ouster builds is to give
04:25machines enough context that they can do things safely while being productive in the real world.
04:32No one wants a world where it becomes less safe with time because we're adopting automation.
04:39So ouster's job at our core is to provide our customers with the safety layer through these
04:45advanced perception sensors like LIDAR to make that all happen.
04:49Yeah, that safety component is just so important and you're completely right that no one is going to
04:54want to deal with any of these autonomous machines if they don't feel safe around them. So with all of
05:01that and talking about this physical AI component of everything, how do you define that term and why is
05:06that so important as we move forward? Yeah, so physical AI means a lot of different things and
05:13actually a core to ouster's strategy and one of the reasons we've had such great financial results,
05:1813 straight quarters of revenue growth, is that we've taken a diverse strategy to our product portfolio,
05:27to our customer base and really we're trying to embrace the entire world of physical AI with the
05:34products that we sell. So we operate in industrial, we operate in automotive space, we operate in
05:40smart infrastructure, so fixed infrastructure like traffic technologies, changing the lights and also
05:46in the coming wave of robotic sensors, so drones and last mile delivery robotics. So we have over 10,000
05:54customers across those four verticals and physical AI is the application of AI to all of those moving
06:01objects and that fixed infrastructure that needs to understand the world around them. So it really is
06:07an umbrella term, but what we're focused on is making sure that we're building practical, deployable,
06:17scalable and safe physical AI. So it's one thing to talk about it and build prototypes, but ouster has made
06:23a name for building the dependable, safe versions of the physical AI perception solutions that actually get it
06:30out in the physical world. So we've shipped over 250,000 lidar and camera sensors in our history,
06:36over 10,000 customers across all these domains. And so we're literally building this frontier of physical
06:44AI across thousands of use cases. You are and it really is going across every different sector,
06:51it seems like you've talked about farming, you've talked about manufacturing, what industries do you
06:55think will see the biggest impact from advanced lidar technology over the next few years?
07:01Well, I think that there's a major opportunity in industrial machinery, whether it's agriculture,
07:07mining, construction, heavy equipment. These are areas where there's a need for automation. It's hard
07:15to find trained operators of this heavy equipment. There's also a need for a better efficiency and safety in
07:24these environments. So a lot of invested environments in industrial are not inherently safe. You know,
07:30manufacturing floors with heavy machinery and things like that. If we want to have more automation,
07:36we also need to make it flexible operating on the manufacturing floor next to humans. And there you
07:43need a layer of perception and safety that ouster can provide. So I think whether we're talking
07:50outdoor heavy, heavy industry, I think there's going to be a huge impact there. I think in industrial
07:55floors, automated machines that are bringing parts to assembly lines more efficiently and doing
08:02productive work on assembly lines, big, big explosion in investment there. And then warehousing and logistics
08:08and last mile. So literally getting goods from a manufacturing site to your doorstep,
08:14that logistics chain is being automated. People are buying more and more things and getting them
08:20delivered to their doorstep than ever before. And that requires increased efficiency and automation
08:26in order to do that efficiently. And this just keeps going. I mean, like drones delivering packages to you
08:33is kind of the end state of this. It's going to take a long time, but it's going to happen.
08:38And the key is to do it safely.
08:40Yeah, I was going to say with all of those safety challenges and all of these new technologies that
08:45you're developing and employing, what has been the biggest challenge as you try to stitch all of this
08:51together and create these safe and yet effective and efficient autonomous machines?
08:56Well, the challenge is really taking this technology and this promise out of the prototype stage
09:03and the R&D stage and bringing it into mass deployed deployments around the world in a diversity of
09:12environments, tackling all the edge cases that you run into in the real world and making it safe in the
09:20process. So Ouster, we work with our customers. We have a lot of, you know, our customers span some of
09:26the
09:26biggest companies in the world that have incredible sophistication and understand their end markets
09:32extremely well. Companies like Komatsu, Google, but we also work with thousands of small companies
09:39that are really learning how to build this technology and bring it to market. And we want to make sure
09:45that we're good stewards educating every type of customer about how to best use physical AI technology
09:51technology in a safe and deployable and scalable way and educate them almost about the edge cases that
10:00come up when you're trying to build a last mile delivery robot that needs to navigate puddles in a rainstorm
10:06in the southern United States, but also the baking sun of the Arizona high desert and everything in between.
10:14So the challenge is really about making these solutions robust enough and safe enough that they
10:21get accepted by society at large so that, you know, we all feel comfortable with some automation,
10:27some level of automation in our daily lives. And we'll definitely have to do that as automation
10:33in our daily lives becomes ever more prevalent. So thank you so much for talking through all of this
10:37with us, Anglis Pakala, co-founder and CEO of Outster Inc. We really appreciate you being here today.
Comments