How AI is Teaching Cities to See

Nick Earle: [00:00:00] Hi, this is Nick Earle, and this is the IoT and AI Leaders podcast with someone who’s been on the pod before, David Ly. He’s the chairman and CEO of Iveda. They’re in the video surveillance industry, and they are tackling some huge societal issues. What you’re gonna hear is how AI has really collapsed the training time for video models, uh, to be able to, uh, identify things as they’re just starting.

David talks about, you know, uh, these e-bikes and, and the problem of e-bikes in cities, and the fact there’s lots of them, and there’s all these regulations, and no one’s taking any notice of them. We talk about litter, we talk about forest fires, early detection, we talk about spontaneous, fires, plastics in the ocean.

This is a company that’s really doing some, uh, very cool things, uh, using the power of AI to deliver [00:01:00] a business case, not a technology case, but a, uh, business case, uh, to solve things that are frankly a problem, but there’s no way of addressing them under the, uh, current models. David’s a really interesting guy.

We start off the pod, with his background from, fleeing Vietnam as a baby, just after the fall of Saigon, and then going over to the US and starting, his company, which is Iveda. So I think you’re really gonna enjoy this. Lots of great case studies, lots of lessons learned about adoption of AI in particular, and how a mental shift is needed, by looking at the positive side of what’s possible.

And, I think we’re probably gonna be doing another one with, with David in a couple of years’ time, to see, what has happened since, because it’s moving just so quickly. So with that, let’s look at, uh, uh, the, uh, second time around with, with David Ly, who is the chairman and CEO of Iveda. Uh, enjoy

So [00:02:00] David, hello and welcome once again, to the, IoT and AI Leaders podcast

David Ly: Thank you so much, Nick, for having me. Thank you.

Nick Earle: Yeah, and, and we were just chatting before we hit record there. Um, uh, we were both surprised that it was actually late 2023, when you first appeared on the, show. And if we got that date right, that means it’s getting on for three years, which is quite amazing. Before we get into sort of what’s changed since, because you’re in a very fast-moving area

I just wanted to remind people, because not everybody listens to every episode, there will be a lot of new listeners. You’ve got a great backstory. You are the chairman, founder, and CEO of Iveda, and I do want you to describe the company.

But before you do, would you mind telling your, your personal backstory? Because you are the only guest that we have ever had that actually describes themself as a boat baby, and, having [00:03:00]escaped from a country, and then a refugee in another country, and then made it to the US. So maybe if you don’t mind, just a, a little recap of your personal story, which is very impressive.

David Ly: Wow. Well, al- always great having you, um, invite me on, on your show. Uh, but Nick, yeah, it’s, it’s been a story for sure. I’ve been, given many opportunities and honor to share my, my childhood story. I am Vietnamese born. I was born in 1975, just shortly after the fall of Saigon. You know, Saigon fell to communism in April 1975.

I was born in November. So, Those few months, the country flipped. So I was South Vietnam. My dad was South Vietnam/American, military, so he had to run, run naked and hide in a jungle, or else he’d be caught and all his, friends and, and colleagues would be caught in the, a reeducation camp.

So as a kid, [00:04:00] your rest of your family was finding their way out. We called it escape. From that point all the way to, I believe, in the early ’80s, all the way up to, like, e- 1984, there were boats leaving, escaping Vietnam into international waters in hopes of getting rescued. You either died or you got rescued.

Those were the only options, and it’s mere luck of the draw. I was very fortunate. We did the same. I was little, approaching four years old, and my auntie, my grandfather bought us passage to escape on a fishing boat with, like, 60, 70 other people. In the dark of night, we went out to international waters.

And the story was we were rescued by a Norwegian fishing liner that gave us a choice. They were stopping in, Osaka. We either get dropped off there or go, go home with them to Norway. But my auntie, at that time was more comfortable because she had spoken a little bit of Japanese from education.

So here we go. I was there, [00:05:00] and I lived in a refugee camp in Japan for a year, and that was my learn, initial learned language other than baby Vietnamese, right? At the age of five, I was … I immigrated to the United States with another aunt’s sponsorship, who was already there.

And I was fortunate to have a childhood, starting with kindergarten in the United States, and grew up in, the Bay Area, California, went to school in San Francisco, and the rest is history. But yes- The rest is history … I’m a Vietnamese, Vietnamese boat baby is the term.

Nick Earle: And a, and a very, I would say typical, maybe it’s not so typical, but it- it’s certainly familiar sort of American success story of a lot of people who’ve, who’ve come from abroad and made it.

So I thought it was well worth you recapping it. So when, when we recorded the first time, you were talking about setting up, Eyevade, really in the video, surveillance [00:06:00] company, space, video s- space, and you were talking about how, you know, video used to be, essentially people in front of screens staring at screens, getting bored, trying to see if, if something unusual was happening.

And often after staring, I remember you saying, when people stare at a screen for a video for a long time, you know, anything, s- big things can happen, and they just miss it ’cause they’re just so gagged out- Yeah … after staring at the screen. So the idea of introducing software, leveraging customers’ existing, infrastructure, the cameras doing edge, edge processing.

And, really it’s the opportunity for cities. And, and one of the phrases you used is that cities’ primary responsibility should be to deliver safety, security, and convenience. And, and video, video analytics, was one of the key ways in which they could do that. I, I think and I, I, I frankly I know the story has moved on.

Certainly the technology has moved on hugely since that, [00:07:00] that time, and that was a much bigger focus. We weren’t even … I don’t even think we mentioned AI. So maybe let’s, let’s, start there. Maybe you, can bring us up to date as what’s changed since then, and then what sort of company, Eyeveda is, is now as a result of the opportunities and the, the advancement in the technology, particularly AI?

David Ly: Yeah, Nick, the AI evolution has certain been, been, very fast-paced. You know, you and I started talking three years ago. Though AI is not new to Iveda even as of three years ago, we started using machine learning algorithms in 2014, evolved to, you know, convolutional neural networks in, in 2018 to launch our first, commercial platform.

But even when I was first speaking with you, we were touting how fast we were, how great we were in accuracy and face detection, license plate, weapons, you know, all the hot topic of the days. [00:08:00] Frankly, three years later, I even tout- I was even touting on how fast we could train certain specific objects and models, for our partners and customers so that they can get to, get in action real quick.

Where we are today, how fast things are, is we’re- we’ve just launched zero-shot training or zero-shot AI, meaning zero training. You know, i- in the past, we’re, we’re, needing to take some video footage, splice it to, into pertinent identified objects by the number counts of hundreds to thousands before it can be trained and, and accurately identified or detected for, for a decent alert, which all still happens today.

Which all still happens today, but the beauty of where we sit today is cities and organizations, Nick, as we talked about, what does safety mean, what does efficiency mean, and what does convenience mean? Well, it’s different. It means something [00:09:00] different to every operator. The beauty of AI video today in our industry is that Iveda is offering it much quicker.

Let’s say you, you don’t want e-bikes, or e-bikes are prohibited in certain sections of, building corridors or even a street. But, how do you effectively regulate or monitor that? You know, police don’t have time for e-bikes prohi- pro- prohibited, to monitor that. They got bigger issues to deal with.

That’s where AI comes in. There’s cameras e- existence in private enterprises and public sectors already. But there, you’re right, till this day, nothing has changed. People are still watching. That’s if they’re watching. Yes, it’s being recorded. Even brand doesn’t matter now. Systems are, are cap- highly capable across the board.

But to make things effective, I mentioned e-bikes as I’m working actively now. I’m sitting in a hotel in New Jersey, but actively [00:10:00] working with, the city of New York. Um, you know, they have cameras everywhere. You and I spoke just like London, but how do you make them more effective? We’re actually gonna look for e-bikes now.

And then when you come to e-bikes, there’s multitude of varieties of how they look, the size and even some makeshift stuff. You know, guys that don’t have a lot of money to spend the $3,000, but they still gotta deliver people’s Starbucks, right? Or In-N-Out. So they make e- electric scooters that, that transport around the city, which are illegal Now AI at an instant can be taught the multitude and variations of, of the difference of e-bikes and send immediate alerts to local authorities that are, that are responsible for, for checking those, those events out, right?

Preventing those, uh, engagements or interaction out there in the streets. So I share [00:11:00] all this with you just to give folks the color of what AI can do, what it means, right? It’s not just a bunch of more noise, agentic AI this, agentic AI that. I always like to take things down to the street level. So yes, if you have a camera and you have, you know, parking prohibited by loading and offloading, trucks only, but- Yeah

normal vehicles are not. Well, how do you watch that? How do you monitor that? I mean, frankly, I park 10 minutes waiting for people illegally. Well, now I’m gonna get a ticket. Now I’m gonna get a ticket because I’m not a truck offloading anything. So these are the city efficiency generating city revenue, right?

That’s a positive. Why are cities investing in this? Yeah. Because they’re gonna give me a ticket for parking at an o- unloading and offloading zone only. I’m gonna get a ticket, 36 bucks to the city, where they lost so many in cars that they couldn’t ticket [00:12:00] before.

Nick Earle: And, and, and, uh, well, two points there. Well, firstly, just that last comment that you made. You know, the fact is that if somebody is parked in a, in a no parking zone or a tow zone, or there are illegal bikes and, and yeah, we have that huge issue in, in London.

And by the way, the biggest issue we have on, on these bikes, and it’s not just electric bikes, it’s, it’s the, like the L- what are they called over here? The Lime bikes that you just, you know, you rent by the hour. Mm-hmm. They, they don’t take them back to the where they, they should take them. They just drop them.

And so, you know, you, you come out onto the street and there are bikes lying all over the sidewalk, over the pavement- Yeah … on their side- Yeah … fallen over, which is bad for, you know, I mean, like blind people just trip over them, and it’s bad for everyone. So actually the city works more efficiently. It’s not the, the fine.

It’s the, it’s the fact that the city flows more efficiently. But I wanted to come back to what you said, because I think that’s interesting, Well, as I understood what you said, you’re [00:13:00] kind of doing the same thing, but you’ve c- but AI has now collapsed the need, I think you said, for hundreds or even thousands of training events.

So previously, I guess you’d have to tr- train, train, train, train. This is a bike. This is another version of a bike. Do it many, many times. Okay, now I’ve got it. Now go and look, look for bikes or whatever. But what you’re… I think what you’re saying is that now you can just say, “Spot, find bikes,” and even if they’re all different sizes and, and shapes and whatever, AI now, I think you called it zero shot, which, which is y- I guess m- is- refers to this fact that you don’t need as much time to train it so you can deploy it quicker.

Have, have I got that right?

David Ly: Yeah, very, very accurate, Nick. And, and the other, you know, we can segue into, natural lang- natural language, search as well. You know, in the past, the AI was leveraging models, as we still do today. I don’t wanna say the [00:14:00] past, right? There, there’s, tech that is still in play- A lot of legacy.

Yeah, yeah … still e- e- still evolving properly and still developing and for greater accuracy. But yes, the reliance on actually a trained model is not required anymore because you and I, the AI can look at a scene now, Nick, and you can say, “Find person riding bicycle.” Riding a bicycle, not… In the past we were person plus bicycle next to, but now the person has to be riding the bicycle, a red bicycle that is, then ID that, alert on that.

So yeah, natural language. Person in blue hat holding french fries. Yes, french fries. We can recognize that. In the past, do you know how long it would take to train red hat, french fries? The, the nuances like Nike shoes, camo pants, that’s natural language that now can [00:15:00] be applied immediately, and the AI is gonna go out there and do its best to find that for us

Nick Earle: and that’s produced this huge productivity boost, I guess, because now you can actually, you can find people, find things, find scenarios, find bikes, find whatever it is, or events using natural language, and that’s a significant breakthrough

David Ly: It’s no longer the, the following the correct syntax.

It’s true natural language, and, behavior is another aspect of detection now that was not possible, three years ago, right? I can actually look for suspicious person in red. That’s natural language, and what does suspicious mean? How do you train suspicious? That’s now, vision language is, is within the machines now, so it’s getting very sophisticated.

I, I believe that, you know, I always [00:16:00] say, at least in AI video, you know, my world, I am- I’ve been u- I’m using a new term. I say, “A, our AI is like a toddler.” Toddlers understand you. They understand your art language. They know mommy and daddy, and they choose to listen or not at times, but they actually know what’s up.

They know what you’re saying. They know what to look for, but they, they still need to learn a little bit for accuracy, right? And as they grow up, they go to school, elementary, middle, high school, college, and eventually get their master’s and PhD. I describe that as evolution for non-tech folks where AI is.

It, it’s not, it’s not a miracle at this point. It still needs to evolve, but at this time, I believe that working with toddlers is better than working with machines

Nick Earle: So if you can, just go a bit deeper on that. If you can actually say, l- let’s look at theft [00:17:00] from retail outlets. Sure.

Huge issue all over the world, certainly here in the UK. A lot of shoplifting. I mean, they… Stores are losing a ton of money. They can’t afford… and what they said as well, “We can’t afford new cameras. We can’t afford, human security guards.” You’re not describing that world. You’re, you’re describing,

I mean, ideally, the quality of the camera really helps, but you’re describing, you know, is somebody acting suspiciously in the store? So presumably, there’s a, there’s an ROI on this technology. Does it scale down to, like, you know, 7-Elevens or gas stations? Or is it still at the city level to be able to, do this sort of thing?

David Ly: Actually, with, with edge deployment capability, it can scale down to a per on-prem basis. Right. It’s a very good question, Nick, and, and the good news to this, it’s all about process efficiency, process and compute [00:18:00] efficiency. We, as developers of, of the soft tech, always aim towards how much can we perform on current physical infrastructure?

So it’s always compute capacity. The good news for all consumers and enterprise is it’s only gonna get more efficient. Yes. It’s you can do more with less power, right? So that, Frankly, you, you described something pretty cool, the retail. You, we, we have tested and successfully found shoplifting. The act of shoplifting is, is highly recognized now.

If you’re loitering around for X amount of seconds, you’re picking up an item, and all of a sudden it goes into your pocket, your purse, or your bag, you’re in a retail store. You know, the AI knows you haven’t approached the register. You have not approached any action that looks like a payment transaction.

Yeah. So you are now shoplifting in the real [00:19:00] world, and that- is the ROI for a lot of retailers.

Nick Earle: And, and I guess it can tell the difference between a bottle of water and a bottle of vodka, because they are, you know, putting a bottle of water in your pocket is one thing.

Putting a, putting a bottle of, a bottle of vodka in your pocket, but the ability to actually spot the, spot the difference, I guess, is also part of it.

David Ly: I believe it was you and I who talked about that. Bottle of water versus bottle of vodka are two different prices, but it, they’re still just bottles being scanned.

You can’t get away with that anymore. It can recognize the Stoli over, the plain water bottle.

Nick Earle: I wanna get back to, some more use cases because the more you think about these capabilities, there’s no shortage of, well, could it do this and could it do that? Well, well, it could do this.

And, uh, and maybe, uh, we’ll even go on to back to your toddler growing up analogy. Is at some point does it start to say, “This is what I think”? Does it start to have its own personality and make its own decisions? [00:20:00] But let’s, let’s just put a pin in that, um, uh, for the moment. Um, given that there’s a lot of cameras around right now, and there’s a lot of edge devices, around, and edge devices prices are pretty cheap, there’s a lot of connectivity around, there’s 5G on the cellular side, there’s fiber, around with, fixed wireless access.

What is the biggest inhibitor, in adoption of all of this? ‘Cause it’s so compelling, David. Is it, is it awareness education? what… I mean, you’re in the business of, of selling this and implementing this, to companies. What, what’s the biggest inhibitor on adoption right now?

David Ly: Nick, you pinned it. Awareness and education. People hear about it, but they don’t hear The proper things about it. What people hear about it are on their, on the [00:21:00] internet, on Instagram, on TikTok, on the news, and you know how news, they give the highlights. They give the words that call people’s attention without any definition or true meaning, you know?

So unfortunately, the mass market- The adoption rate is not … It, it’s picked up. It’s picked up. That’s why you asked me, “How are you doing?” I go, “You know, could be better. I, I’m excited about business.” Yeah. But-

Nick Earle: Yeah …

David Ly: general adoption is, i- or the lack of, is due to lack of education and true awareness of how easy and available it is today, Nick.

People still, without the proper awareness, believe that it’s something out there. It’s too expensive. It’s in the movies. I don’t know. It might be too complicated for my organization. I’m in these meetings all the time, and it’s the same hurdles that people have to come over. This is why I’m big in speaking what I call [00:22:00] street-level language.

I just get down to street-level talk with, with the highest levels of contacts and organization, Nick. You get into convolutional neural, language and neural networks and transformers, and who cares? Who cares how- Yeah, yeah, yeah … you built it? You know? Show me how it’s gonna work for me today, and that’s where the majority of our markets don’t have the opportunity.

These tech guys go in and start blurbing out they’re the best thing out there next to sliced bread. It’s … I’m happy they’re doing that ’cause they’re turning business to companies like Iveda now, because we speak real English, true English, in describing what the heck the tech is gonna do for me today.

That’s key. But yes, that’s why I like doing these shows with you, is I get the chance to speak street-level language. I’m not here to spell out for you how we build AI. I mean, you can go to ChatGPT. It’ll tell you all you need to know. I [00:23:00] don’t need to do that. You and I, as, as pros, need to tell our audience how they can get started, how they should try it, feel it, and then put a budget to it, and then it’s not so scary.

Not so scary.

Nick Earle: And, and, and it, it … No, and, and, and, and actually what you’re describing is the, was, was the, basis of creating this pod, over five years ago. And when we look at our data, and forgive me any listeners or viewers who heard this before saying, “Oh, he’s gonna tell this story again,” but we are proud of it.

Two-thirds of our list- We have listeners in over 100 countries, but two-thirds of our listeners are director or above. So b- business-level decision makers with budget, which is unusual for a, in quotes, “technology pod.” One thing we d- we never talk about on the pod is the tech, because exactly to your point, we talk about the business case and what the ROI is.

Not the tech. Whereas a lot of, [00:24:00] IoT or AI, definitely AI, podcasts and shows get, you know, deep into the tech and the latest ChatGPT model and it really is a geek fest. And I think that’s part of it because as you say, if you’re a store owner or a gas station, owner or a branch manager or, or even, a business, a, a, a business executive in a, in a medium to large company, it’s not the tech ’cause the tech…

By the time you look at the tech and you study the tech, it’s moving so fast. You know, we talked about that, didn’t we? You know, I do some lecturing at the local university here in the UK, and I was trying to explain to the students just how fast the tech is moving and why they needed to concentrate on soft skills like critical analysis, judgment, communication skills as opposed to the tech.

Because, you know, Moore’s law, driven the industry for 40 years is, is 40% a year, so, so everything gets better at a rate of 1.4, [00:25:00] 1.4, 1.4, 1.4, and that’s done that for 40 years. But actually, based on a, a, a, a basket of benchmarks, AI is, is getting… Its capabilities are getting better at four, so 1, 4, 16, 32, 64.

I was speaking to a business school, class of, a few hundred graduates and I said, “Look, you’re first year students. You’re gonna graduate in three years’ time. Um, uh, this thing is, is… By the time you graduate and you’re entering the workforce, you think AI is powerful now, it’s gonna be 64 times, more powerful.

So… and everything that you learn about AI now will be completely out of date and superseded, and it will seem like, old-fashioned programming languages that we’ve forgotten and techniques and things like that.” But it’s the business case is, and I agree with you, is the, holding back the adoption.

Which I guess leads us to the fact that that means that the decision makers I would guess that you [00:26:00] target And I think you said that, are not the… You’re not selling to the technical audience, you’re selling to the business audience. When you, when you talk about street language, you’re, you’re getting to the person with the business problem and, and showing them c- case studies and, and what it can do for them.

Because you deliver the whole thing, right? You d- you don’t give them a, a, a set of tools and say, “Get your IT department to connect this to that, and tune this, and, and, and y- you… And put these prompts in.” You deliver a managed service that actually allows them to use natural language to, to, to deliver a business outcome.

That, that’s kind of the way you, you go to market, right?

David Ly: That’s very accurate, Nick. We don’t speak the tech talk, right? We don’t geek out with organizations because that’s the quickest way to lose money

Nick Earle: Geeking out. We got street talk.

David Ly: We

Nick Earle: got geeking out. I love it.

David Ly: Yeah. No, I, I’m- Yeah … it’s, I’m a simple language guy, you know? You can trust- Yeah … that I’ve got all the vocabulary behind me, but I’m [00:27:00] not here to, to, toss that around. I, I like people to understand and comprehend clearly what we are delivering. And Nick, you’re so right. we all run and support businesses.

What we really wanna do is understand the cases, and the sell starts with speaking with folks who have problems and issues that they need to solve and overcome. It’s not about pitching the tech into the organization. Once you can prove that you can do something, resolve an issue, or help generate more results or money, then that leader, right, or decision maker is then going to be your advocate in the tech side, kindly walks you back into the tech room and say, “Ladies and gentlemen, this is what we need to, to get and accomplish something we’ve always been looking for.”

And what is that need? The IT server integration, right? This is when IT has a [00:28:00] very clear mission and directive that they understand. They’re not just, “Oh, God, another piece of tech we have to manage. Ugh.” I hate that, and I’m, I feel that for them, too, because nobody gives them a, a reason or a purpose. It’s just directed as, “Hey, add this tech into there and learn it and deploy it.”

No, it’s different with us. We built and designed technology that is seamless to IT. What’s another server on your rack, right? You, you spent 10 years in your career managing servers on a rack. You’re doing the same thing. No heavy lifting. Multiple servers on a rack. Clustering. No heavy lifting.

Integration with clouds. No heavy lift. Everybody’s already experienced. You- Then the software, so turnkey that IT installs it, provides access to it, authentication credentials, just like they light up a new laptop for a new employee. [00:29:00] And then the operator side of the business gets to log in and apply it instantly, Nick.

Nick Earle: Well, I, I think I know what you’re referring to. I, I mean, I remember, you talked about how long, when you were born.

I was born before you, and I remember, even if IT had the tools, they gave you permission to access the system. And if you asked for a new application, they had to get storage, they had to get compute, they had to get- Yes … the database, they had to get the stack and the…

And it could take five, six months before they, they gave you permission to access their system. And this- Yes … that’s completely changed now, the other way around. You know, you mentioned you’re in, I think you said New Jersey, New York area. So I, I… We talked about security and retail and video surveillance and, not having to train the models like you used to.

I believe you’re doing some work with, the New York Fire Department. Yes … and that’s int- that’s an interesting, area. Maybe you could just, expand on that a little bit to give another example of a, a different segment where this would [00:30:00] apply.

David Ly: Yeah, no. We’re, we’re proud to share this, and, and through another strategic partner that helps takes our m- our technology to market, we’ve been able to earn an opportunity.

It started about two years ago, Nick, so this is not new. It’s evolving, which is fantastic, news and opportunity, is we started with f- smoke and fire detection, right? The a- the agency already had deployed remote cameras throughout the, the, the marshes and dry lands, wetlands and dry lands that change.

there’s… We heard spontaneous combustion. We hear about that in school, but in the real world it actually exists, and I wouldn’t think it exists in certain areas that I’m learning about today. At nighttime in New York at, uh, you know, brushes near the water, can you believe that it just –?

Starts lighting up. Who would think in the middle of the night, brushes that are near [00:31:00] water would- What, that- … spontaneous combust?

Nick Earle: And before we talk about your, how you can address that, just help me understand. Why do they do that? Is that the heat or people throwing cigarettes or- It- … or do they not know or?

David Ly: No. S- see, the, what I’ve learned is the temperature change. Pressure and temperature change in, in that part- in these particular areas, and you can’t, you can’t predict them. You just can’t predict them. No, and no, people are… This is way too far for somebody to toss a cigarette and you, you know, to cause that.

So baffles me as well. So this is why they place remote battery-powered cameras out there and, solar powered-

Nick Earle: Like trail cameras- You can pick this up … for animals and things like that? Yeah.

David Ly: And now, in the past was, is always after the fact. Okay, it happened, let’s review footage and see what’s going on, but there, it’s already damaged, et cetera.

We’ve been able to apply smoke and fire detection that on the early detection of smoke, an alert’s already [00:32:00] provided, and then authorities would take actions under their, their protocols, and that’s expanded all the way to, city cameras, agency cameras for smoke and fire of, you know, anywhere between buildings, et cetera, just for early detection More recently, we’ve been a part of the, uh, Sail forth 250, America’s 250 birthday.

Uh, last 4th of July was all international boats and ships were visiting the, the Hudson, and we were part of vessel detection. You know, anything from roll-on/roll-offs, sailboats, jet skis, military vessels, ferries. Those are the things that in the past would take forever to train. We were able to prepare the model over Memorial Day weekend in prep for July 4th weekend here in the United States.

And now, you know, I mentioned, e- e-bikes. Same thing. The [00:33:00] same agency, same city, has expanded the utilization of AI to now monitor these prohibited activities. So from smoke and fire detection to prohibited e-bikes on a, on certain areas of the street, these are the efficiencies and applications that apply to, to major cities across the world.

It’s not… When people hear AI, it’s not always the, the movie stuff, it’s the real world problems and issues that we face today, similar to you sharing people just dropping the bikes. They’re a nuisance, you know? Yeah. And a safety hazard. And fr- frankly, it makes your city look dirty.

Nick Earle: Oh, it’s terrible. I mean, L- London- Yeah

B- my wife and I just go crazy. We have an apartment in London, in Kensington, which is a nice part of London, and we come out and, the pavement, the sidewalk, is littered with green bikes lying on their side. Right. I don’t… Why the hell do we allow this to happen? And it’s a logistics problem, they claim.

Well, I think there’s two problems. One is they claim that they don’t know where the [00:34:00] bikes are, which I kind of find hard, because there’s technology in the bike which identifies where it is. But also, they’re not following up and finding people. They’re not charging them for not going back, which has to be a process problem.

It must cost them too much. Um, there must be some reason why they don’t fine. But if you could, as you say, I mean, instantly think about a case study. If you can identify that here’s a bike, in this case it’s a company called Lime.

David Ly: Piles

Nick Earle: of them. Someone’s just dr- Yeah, it’s someone’s dropped it on the sidewalk there.

It’s nowhere near where they’re supposed to take it back. Bing. They know who the person is, they know who, they know where the bike is. They could debit their credit card. Let me tell you, it would, it would stop.

David Ly: Stop

Nick Earle: immediately. It would stop pretty- Yes. It would stop immediately- Yes … because there’s no punishment.

David Ly: Gotta hit ’em in the wallet.

Nick Earle: Then suddenly, amazingly, it stops. I saw in the notes as well that you, you’re doing some even bigger stuff, if I’ve got this right Looking at, environmental initiatives like, marine life and plastics in, in, in [00:35:00] water, which is another really big issue.

Maybe you could just talk a l- a little bit a- about that

David Ly: Wow, you keep good notes, Nick. Your team does well. I, yeah, you know, I, I touched on… Because we talked on zero training, right? Zero shot. We’re working with another partner that works with NOAA here in the United States, the Nash- National Oceanic, Atmospheric, A- Administration, where we monitor ocean life, weather, all that.

But a lot of it is really tracking, pollu- pollutants in our ocean, and a lot of that results in garbage, trash, illegal net dumping, all kinds of stuff. And how do you identify that? Humans, eyeballing from ships, sometimes detailed satellite images, but it requires a lot of manual review, manual investigation.

So we’re working with the agency [00:36:00] and our partners to leverage video, where video can more rapidly identify and alert on the potential of such events, you know, nuances occurring. We’re even looking at radar patterns. What I mean by… We’re not ingesting any radar data. We don’t know how to do that.

But what I’m saying is, remember the radar screen does appear on a screen into a video- Yeah … format of some manner. We take that video display, and then we run it through like a camera, Nick. So now the AI is gonna look at variants and dots and, and, and-

Nick Earle: Yeah, yeah … elements- of

David Ly: the radar.

Nick Earle: Seeing an X-ray.

David Ly: Yeah … through professional guidance of the expert at hand, he’s gonna say, “Oh, that kind of buildup represents A, and this kind of buildup represents this.” We’re now watching radar development via AI

we’re doing real-time [00:37:00] alerts on potential formation of radar imagery. Pretty cool, huh?

Nick Earle: That is pretty cool. Sort of the marine version of dumping the line bike in Kensington. I like that. And you think of littering in general, and people tossing, litter out of cars and… I mean, all of these things.

I mean, I guess we all hope that in years to come, AI, which sometimes gets a bad rap, and people are concerned about it. I was listening here, driving here today, about the big pushback in the US and the elections on people who have supported building local data centers, and people saying, “We don’t want AI data centers.

It’s gonna take our jobs away and not gonna recruit anybody locally.” But the, the, the potential for AI to do good- I mean, that’s really what, what your story is. These are major, major issues that could make the world a much better place, which is why I asked you about the, about the adoption. We’ve got some great case studies, and maybe we’ll do a, you know, a third, part.

Hopefully not in, two and a [00:38:00] half or whatever it is years’ time ’cause in two and a half years’ time, I think the tech will probably be… AI will probably be 20, 25, 30 times better. Put the adoption issues to one side. At some point, there’ll be enough critical mass, and there’ll be enough big, visible projects, whether it be, sending, first responders to a fire.

I was out in Sedona, last week… You’ll know that well. It’s in your neck of the woods … on vacation, and they had this big what’s called the Pocket Fire out there, which was, had been burning for, like, 18 days. It probably still is burning, and the helicopters going overhead. I mean, the number of, you know, thousands of responders trying to put it out ’cause it had been burning for a while because it was in a remote canyon before they discovered it.

E- every case study has huge societal benefits, being able to react quicker and make the environment, the city, the, citizen experience or whatever better and better and better. But, but where, where do you think it would go if we time traveled [00:39:00] forward? Let’s say it’s now 2028, so let’s just say two years’ time, and we’re doing podcast number three, and we get through the, “Oh, good to see you again.

What’s happened since?” What sort of case studies do you think/hope you might be involved in, or will it simply be more of implementing… I mean, I think you wanna implement more of what you’ve got these, these examples of ’cause the opportunity’s so big, but can you predict or hazard a guess of the types of case studies that we could potentially be talking about in two years’ time based on what you see happening out there?

David Ly: Yeah. Great question, and I… In two years, I think, Calling on all the experience that, that I have I’m gonna keep it very real for you, Nick. In two years, [00:40:00] I truly feel that you and I will be talking about customer experiences, because it’s gonna take some time. I, I’d like to share some futuristic movie-like prediction for you, but adoption is going to increase, which is for the better.

But I truly believe, knowing what I know and the experiences that I’ve had the opportunity to, to have had, is organizations and even governments, it’s not like the dystopian movies. People are looking to improve human experiences, making things faster, more convenient, cleaner, and better for all of us.

I look forward to sharing with you what some of our organizations have accomplished. How are you and I not getting pissed off at the p- plane, w- getting on a plane anymore? How you and I are n- less stressed crossing a bridge, from one city to another? How, how [00:41:00] you and I are improving in our mental health because we’re not dealing with some of the crap that we’re seeing, that we must every day in our lives?

I believe that we’re beginning … We’re gonna begin to talk about human experiences that actually differ for the better because of AI, and- Yes … and how … Yeah, and I’ll leave it at that. I think, I think if you and I talk again closer to two years from now, that will be it, right? Yes, technology will evolve. Yes, there’s gonna be new techniques and methods of compute and processing.

But frankly, how has it changed you walking out your front door at home grabbing a, a coffee, right, Nick? That’s … I see- Yeah … that as something cool to talk about. That’s when we know- technology’s

Nick Earle: I, I think right now we still have a suspicion out there, the surveillance, the privacy, e- examples of where it goes wrong on a small scale get amplified like, like crazy on, you know, facial recognition, it can [00:42:00] make mistakes, and the media tends to focus on that. But i- we don’t give anywhere near enough attention to, to the good it can do on a

The case studies that we’ve talked about is, is doing good on a huge scale. And yeah, if in two years’ time we can have a mind shift that A- AI is an enabler of tremendous things, other podcasts we’ve talked about medical breakthroughs, disease. A whole bunch of, physics, chemistry.

certainly in terms of, like you say, walking out the door to get a Starbucks although it might arrive by drone, by the way. but, you know, the, if the average person in the street, as we used to say in London, the man on the Clapham O- Omnibus, if the average person in the street’s life is better as a result of AI and they know it, and therefore they start saying, “Why am I…

why are you not implementing this?” Which they’re not saying today ’cause they don’t know about it, [00:43:00] then the, the world will probably become a better place even quicker. And, yes, we’re not there yet, because, you know, awareness of, of the capabilities of AI is still so tiny amongst the general population.

And the whole debate about, oh, it’s gonna take my job and, and, it’s gonna suck up all the power, and whatever. But, but yeah, if we can get there or at least partially there in two years’ time, I, I think we can then start to make great changes, to a lot of things and solve a lot of problems that, that frankly are still problems, particularly on the environment side and that last case study.

So maybe that’s where we should leave it, David, and, and say, well, let’s check in in a, in a couple of years, and see what other great case studies that, that you’ve got. But in the meantime, if, let me just do the shout-out. If people who are listening to this, and I’m sure they’ve loved it as an episode, if they wanted to find out more about your company, how do, how do they do that?[00:44:00]

David Ly: Please, uh, first follow us and, and stay tuned on all that we’re doing in real time on all the social media that, uh, is out there today. But, uh, visit us at Iveda.com, reach out on our contacts list and say you wanna talk to somebody. I will speak with you personally. It’s gonna be fun. We’ll arrange things, and we’ll, we’ll duke it out, and we’ll speak real world language and see what’s, what’s happening.

Nick Earle: Street talk and not geek talk.

David Ly: Yes, sir.

Nick Earle: Okay, David, always a pleasure. Thanks for being, a guest on the podcast for the second time. I look forward to checking back in with you, again to see how the, development has moved on and whether or not the attitudes in society have also changed as well.

So thanks for being my guest on the, IoT and AI Leaders Podcast.

David Ly: Always happy to speak with you, Nick. Thank you so much. I’d be happy to chat with you again.

Nick Earle: Thank you very much. [00:45:00]

David Ly: Yes, sir.