Expanding Strategic Partnerships in 2024 with Sony Semiconductor Solutions’ Jim Lenox

This is Textron tv. Hey everyone. Welcome back here to Textron tv.

Our next guest is a first time guest here on Textron tv. Uh, his name is Jim Lennox. Jim is the VP of Sony Atri.

Sony Semiconductor Solutions for those who are big tech drunk TV fans. You know, we, we periodically do check in with our friends at I trios to find out what's new and what's going on there. Um, I'm both happy and sad to report that, uh, the spokesperson who usually comes up on our show, mark has retired from Sony Rios and he's, uh, enjoying his life from what we're told.

And I'm happy to introduce, uh, the new spokesperson here, or the new guest on Tech Drunk TV from Atrios. His name is Jim Lenox, as I mentioned. Jim, welcome the text Drunk tv.

Yeah, thank you Alan. And you're right, mark is a hard act to follow, but we love, uh, joining Techstrong tv and thanks for this check in with us. It's great to meet you.

Not a problem. So, Jim, you are the new kid on the block. Give give our audience a sense of, of who you are, what you've done, where you've been, and how'd you wind up here, Alan?

I've never been called the new kid on the block before. The old new kid on the block. You know, it, it feels good though, doesn't it?

It does. It feels great. It made my day.

Thank you for that. Yes, you're welcome. But yes, yes.

Uh, I, I joined, uh, going on eight months ago now, uh, the Sony team. So, uh, I love being here. Um, I actually started my career as an engineer, so I've done a lot of engineering, uh, in the past control systems, digital control systems, et cetera.

But, uh, I went onto the business side of, of this whole industry some years ago. Um, I actually joined VMware when it was a really small company, when it was about 50 people. And I was there many, many, many years.

And I left when it was probably about 25,000 people. Um, and it was, uh, Was that during the Broadcom? Was that like at, at the Broadcom purchase or before?

It was, it was, uh, way before when we were very private, before our first venture funding, frankly, uh, before the EMC acquisition in the spinoff and the delac, all, all, all, all sorts of events that happened. But I was on the business side there and I was running, um, one of the geos, one of the three geos for much of my time there I was in Asia Pacific. Um, and that was great, right?

That was great. And when I left VMware, I joined a couple of the technical leaders who had a technical startup in the machine learning space. Huge.

And that was, uh, the beginning of my journey in the machine learning and AI space, which I've been been really passionate about for a lot of years, and was at a couple of startups, most of them acquired. And I was really, really drawn to what Sony Rios was doing, and that's what led to my journey here. You know, we know that that machine learning AI space is, is really hype now, but it's also got a lot of history, a lot of difficult history.

You know, typically machine learning AI is tough to get working for you, but the approach we took at Sony really appeal to me making things simple, consumable, somewhat elegant, and very Sony fashion. Yeah, I mean, almost by definition the whole machine learning AI space is, is a big data problem, right? You gotta have a sufficient amount of data to, to, to recognize your patterns and, and create, you know, the actionable intelligence you want.

Um, you know, it's funny, I I have a friend of my good friend of mine, John Willis, and John's working on his next book right now, which is basically the history of ai. 'cause you know, everything is AI today, of course. And it's been almost two years, a year and a half, a little over a year and a half since Chad GPT and all of that.

So we all talk ai, but the fact of the matter is, AI itself as a notion is not new. It's been, I mean, John's trace the back of the 1940s event, and, uh, you know, people talking about Turing and, and stuff like this. And so, and when we talk about machine learning and ai, again, it's not, it's not new necessarily.

What we can do is new, right? That we have new tools, new capabilities, new processes, but the, the, the, the ideas, if you will, are not new, right? The, what we think we can do doing it is another story.

Jim, before we jump into that though, or, or as part of jumping into that, look, everyone in our audience knows Sony, right? Worldwide leader in electronics, whether it's TVs or music or other electronic devices, uh, Sony is, is a quality stands for a certain level of quality in the world. I ts and Sony Semiconductor Solutions may be something new to a lot of our audience.

Now, if you wouldn't mind taking a little time and kind of explaining that to us. No, I'd love that. I'd love that.

And I think that deserves some, uh, some discussion. So, so you're right. So, um, I'm from a division of Sony that is Sony Semiconductor Solutions.

So why would you be talking to Sony Semiconductor Solutions about, you know, edge AI and an edge AI platform? Well, the reason's this, so our division of Sony is, you know, the biggest provider of image sensors in the world. And when we're talking about image sensors, we're talking about, you know, those sensors which are behind the lens of your camera.

So probably the cell phone you have in your pocket. Alan, um, has a camera that's, you know, uh, has a key component that is a Sony image sensor in it. Most industrial cameras in the world would have a Sony image sensor in it.

And what we've done is we've taken that highly durable, renowned, highly sensitive, reliable image sensor, and we put on that same little wafer and AI processing capability. Uh, so that's what we've done here to turn, uh, basically that technology into the world's first intelligent image sensor, right? Wow.

And yes, so, and the, and what that allows is, you know, instead of sending just videos and images off of a camera equipped with our chip, we're sending notifications, information, data about what our customers care about and what they care about is all captured in an AI model that runs right on that chip right behind the lens of the camera. I love it. I love it.

And, you know, I don't know, I think people with what you said, I hope they grasp it. I mean, this is, this transcends Apple and Android and, you know, and not just fault cameras, every industrial kind of camera in the world really has, has some of this Sony technology in there. Now, of course, that lends itself to having a lot of partners, right?

Because I mean, there's partners and there's partners of course, right? But every company that's using this chip in their device, in their camera, but is in some, at some level a partner, Right? Yes.

Yes. And, You know, so this is, this I os business is, is a, is a huge biz dev play, if you will. It a huge partnership play here.

Talk to us about that, Jim, and maybe what's in store in 2024 around you. And again, not all partnerships are equal, some are more strategic than others, but talk to us about this. That's right.

And we've talked about some of those in the past. Um, you know, listen, so when you know, when you're sending, uh, data off of a camera, right? And information and notifications, you know, what people care about.

The question is, what do our customers care about? And different customers care about different things. So we have these engagements and these partnerships across the logistics industry, across, uh, across manufacturing industry, um, across retail industry.

And you ask yourself, what do they care about? And they care about a lot of things. For example, in retail, um, we've done this implementation, um, at seven 11, um, 500 convenience stores.

Their retail media team is interested to see, and this is a very typical use case for us in retail. They're interested to see what's going on in the promotional areas. In the case of seven 11, what's going on at the digital science?

Um, what, how many people are being attracted to a given space in the store? Um, how many people are looking and how long are they looking at the ads going through these digital signs? And how are they acting on that as far as what they buy?

So, you know, a lot of momentum happening across different industries. Uh, we can park these AI models, you know, again, to give insights on exactly what you care about, right? And, and send that data and allow you to get kind of better analytics around it.

And this retail is really fun for us, Alan, uh, what we're doing there, because when you think about it, and you know this well, and you talk about this, you know, the online e-commerce activity, you have complete visibility to everything the customer does. What they look at the ads, they look at what they put in their cart, what they take out, what they buy. But think about when you go to a store, what do you see?

You see some point of sale, you see what folks buy. So there's much more attention to behaviors and activities in the store, right? Looking at customer propensity, what's leading and attracting their attention, right?

And when they're spending their activity. So we're really happy what's going on here at seven 11? 'cause we have a lot kind of going on in that sector in particular, um, logistics as well, right?

We've done an implementation at Matsui soko, and this is a large warehousing, large logistics operation. And there, we're doing what we do in many places. We're looking at the yard trucks coming in, um, making sure they're not waiting too long, making sure the receiving staff aren't loitering too long.

You know, triangulating, arriving trucks with their births, making sure our staff and the loads meet each other exactly the right time to minimize waiting time by the receiving team and the truckers as well. So a lot going on as far as use cases and adoption across those industries. It is, Andy, you know, what it, what you mentioned, it scratches the surface, no pun intended.

Scratches the surface of, of, uh, of the possibilities. Here. Let me bring something up to you, Jim.

One of the things we're seeing with the ubiquity of, of this kind of technology and a, and marrying AI to it, privacy concerns. I don't want you to see what I'm doing in the store. I don't want you to see what I'm picking in the store.

I I want to, you know, I wanna keep some stuff private. I, I don't want, it's shared, you know, and especially if you happen to live in an EU country, you have a right to that, right? Um, how do you, how does, how do you deal with that?

Yeah, it's, you're spot on on that, Alan. It's a, it's a huge driver of our business and definitely a differentiation in our approach. You know, we're talking about, you know, a edge AI vision solution, which means all the AI processing is done on the edge, basically in the camera device, on the chip, right?

And when you keep everything on the edge, you know, inside the device on the chip, we don't have to send images and videos with personal identifying information anywhere, right? We basically do that processing there and we send stick figures and boxes. We send coordinates and text data, we send inference data, we send data right about again, what our customers care about.

And in the seven 11 case, it's number of people, you know, a gaze times, locations, et cetera, right? But not sending their personal identifiers. So it's very much a driver.

And when you do this in a logistics setting, you know, I was talking about our use case at Mitsui SoCo and what we do around that, you know, you're, you're talking about, you know, not just privacy issues, but unions and, and employee concerns, HR concerns. Sure. So you're absolutely right.

When you start parking cameras, places, it raises all sorts of antennas on privacy concerns, which we very much address in our offering, sort of at that architectural level, right? Sure. Yeah.

And the other thing you gotta just marvel at is, you know, Jim, you and I have both been around back for a while. Long time. The amount of processing that gets done on that little chip on the camera right, probably is greater than what the Apollo, uh, 11 folks, you know, took to the moon.

Yeah. Uh, it's just, it's phenomenal when we, you know, because, you know, this whole Nvidia and GPU and all of these things, right? 'cause the amount of, of horsepower needed to do a lot of this kind of stuff is, I mean, it's, you know, it, it's phenomenal.

And the fact that we could do it on a small little chair behind the camera lens, It's, it's, you're, you're, you're spot on. You know, like, you know, our, our values as a company have always been about, you know, making things consumable and available, you know, effectively with the image sensor, we democratized, you know, the ability to take pictures and made it widespread. And what we're trying to do here is democratize vision ai, right?

Um, by making it more broadly available and simple. But you're right, right? ai, just what you said, AI normally implies a giant pool of GPUs underneath it.

And we're running this on a little wafer behind the lens of the camera. So there's trade offs, you know, there's clearly trade offs. Well, obviously On the size of the AI model we're running and what we can do, but generally, right, if, uh, you know, being able to send data about what you care about, you know, if that's definable is usually achievable.

But we have, you know, more technical partnerships that are helping us along the way here as well. We have, uh, we just signed a strategic agreement, for example, with plug and play. You know, this is, this is, you know, an organization that incubates and accelerates thousands of startups, right?

So we're, we're really happy about like being able to foster innovation by this sort of agreement with The next generation. Yeah. We may have talked in the past about Raz Pi.

Yeah. You know, we've always had an agreement with Ra Pi where we produce a lot of their units and provide them components and we're with within our business unit, um, extending our reach into their community, uh, by collaborating with them. So there's a lot of exciting things going on, you know, to, uh, to build the apps and the models necessary to serve a lot more customers and use cases.

Love it. Jim, unfortunately, we, we blew past our 15 minutes. I gotta wrap it up for people who want get more information on Nitris.

What's the best website? sony/semicon/uh, slash uh, dot com slash ian for English. You can get it there.

Thank you. That's right to Victoria for that. Jim, thank you, Victoria.

It was a pleasure meeting you. I hope we'll have more conversations like these soon. And, and best of luck.

Thank you, Alan. Really appreciate this. All righty.

Jim Lennox, VP at Sony I, Sony Semiconductor Solutions here on Textron tv. We're gonna take a break. We'll be back.