Physical AI Will Transform 40% of Global GDP | David Gal

David Gal (0:00)

We have the millions of vehicles that are on the road, the millions of cameras that are out there, and those are all just streaming up trillions, literally trillions of data points every single year. That data asset enables us to build models which are very practical.

We can now train a model to detect whether somebody is drowsy, whether eyes are drifting off the road, or maybe whether somebody hasn't parked unsafely on the side of the road or all sorts of things. That iteration, the speed and the ability to deploy the impact is only possible with the data asset that we've spent time building out.

SPEAKER_2 (0:34)

Hello, everyone, and welcome back to MTS. Today, we are joined by David Gal, the VP of Product and Engineering for Connected Equipment at Samsara. And yesterday, Samsara launched their Samsara 360 camera, which it calls the industry's first single mount 360 degree AI camera built for forklifts, excavators, and airport ground equipment.

Thank you for joining. I'm excited to have you. What I really like about Samsara, and we've had some folks from the team on before, is it changes my personal perception about what I understood physical AI to look like. A lot of the conversation right now, we hear humanoids, we hear robotaxis, but this is existing in our existing physical infrastructure. So I would love to get your high level perspective on how you see physical AI playing out for the future.

David Gal (1:30)

Yeah, I mean, first thanks for having me. And then second, physical AI, it's a huge topic right now, obviously. We've been kind of doing this for the better part of a decade.

If you think about what the physical world looks like, it is true that we got robotaxis coming and way more than those things that we're very excited about. There's a enormous, much, much larger population cohort of assets and equipment out there that are not from the OEM factory, ready to go, you know, remote taxi kind of things. And then these job sites and work sites that really do need to benefit tremendously from technology that can help keep those organizations safer and more efficient. So you imagine a job site that could be mine, that could be a construction site, it could be an airport, a warehouse. There's a lot of risk factors to those employees that take place on a day-to-day basis. And so that's where we see physical AI really manifesting right now for us. Not only on the safety side, but the efficiency side. Are those assets being used correctly? Are we burning too much fuel on them, etc.? So this is kind of how we are applying physical AI to the world of physical operations. We've been doing so for the better part of a decade.

SPEAKER_2 (2:37)

And for someone who's maybe not as deep in the weeds here, not as familiar, what's the size of this market? How big is this?

David Gal (2:47)

It's huge. So we think about like 40% of GDP globally, going into the world of physical operations. Samsara's role in this is to help orchestrate those physical operations customers. If you think about these companies and organizations spanning from transportation, warehousing, construction, field services, manufacturing, food and bed distribution, all the way out to other industries, there's just a lot going on there. There's spend on maintenance, there's spend on people, there's spend on fuel, capital, safety. And so this is kind of where we really help these customers out, is driving those costs down, make them just operate a little bit better.

SPEAKER_2 (3:25)

I know you guys are starting off with forklifts, trucks, and all of that other equipment. How do you determine where to begin?

David Gal (3:34)

Yeah, great question. We've been obviously around since around 2016, 17 on this stuff, and we primarily started with vehicles. Most of our customers came to us early on in the days of the ELD mandate and said, help us out with that. And so we built a gateway that we've digitized on millions of vehicles around the globe. They then came back to us and said, hey, can you help us run safer? Can you make a safety camera for us? So we did that around, I believe it was 2017 at the time. We shipped our first AI Dash camera.

And that's really where we start participating. These companies don't just own vehicles though. And so they very quickly started coming to us with other problems. They said, can you help us with our trailers, with our tools, with our yellow iron, etc. And what ended up happening is over the course of time, our tech has obviously improved. We've gotten new sort of technology tailwinds at the same time. And these organizations came to us and said, we love what we've done for our vehicles on safety. Can you please do the same for our equipment? And specifically, some ground service equipment organizations at airports came and said, look, we hit airplanes on those belt loaders. We can come up, we hit, like help us stop that, or we have fatalities on a runway where somebody actually gets run over. This is like real stuff. And so help us.