The €1 Breakthrough: Scaling Smart Labels for Impact

Smart cities are making headlines—but what’s driving real ROI behind the scenes?

In this episode, we explore how consultative IoT strategies, not just new tech, are reshaping urban infrastructure and logistics. The conversation centers on smart labels, AI, and sensor-driven solutions that turn everyday operations into measurable value.

You’ll hear how municipalities and enterprises are rethinking connectivity by starting with ROI—not infrastructure. Key takeaways include:

• Why the one-euro price point is pivotal for active smart labels
• How smart parking systems recover costs in under a month
• The hidden cost drivers IoT can eliminate
• Why AI models are only as good as the IoT data they feed on
• What consultative selling unlocks that tech-first approaches miss

Tune in for a grounded take on what makes IoT work at scale.

Transcript

Intro:
You are listening to IoT Leaders, a podcast from Eseye that shares real IoT stories from the field about digital transformation, lessons learned, success stories, and innovation strategies that work.

Nick Earle: Hi, it's Nick Earle. Welcome to the IoT Leaders Podcast. Today we're going to Romania and we're going to be talking about three subjects, actually. 

One is smart labels, which we've talked about before on the pod. We're going to be talking about the ROI for smart cities and how it really works and that's what municipalities are really doing with IoT. And then towards the end of this one, a very interesting discussion on how AI is totally transforming how people look at IoT and even to the point of saying is the true purpose of IoT a different to what we thought? 

A little while ago. It’s true purpose really is to get the data to train the AI models. And so, my guest who's going to go through all of that is Silviu Neghina, as I say, he's in Bucharest, but he actually works out of Paris. So, he's another one of these crazy guys who flies to work. And also has interesting background, he was playing tennis professionally at a pretty high level when he was younger. So very interesting guy. I think you'll enjoy the conversation with him. He's got a lot of ideas about IoT and what's real and what's not real. And they are pushing the AI models pretty heavily at his company's current company called VizioSense. 

So here it is. Enjoy. Silviu, welcome to the IoT Leaders Podcast.  

Silviu Neghina: Thank you. Thank you for the invitation.  

Nick Earle: Thanks a lot. You're very, very welcome. And you're, I think you're the first guest, uh, that we've had on from Eastern Europe, but I believe you are in Romania, is that right?  

Silviu Neghina: Yes. I'm based in Bucharest in Romania but traveling pretty often to Western European countries where I have my teams and my customers. 

Nick Earle: Okay, so let's, let's first of all delve into you and your background and the reason why you're traveling to Paris a lot, because you have an interesting background from our initial chat before this recording in terms of your experience with the IoT. We'll get into what you're doing now, but maybe you could just give the listeners just a quick thumbnail sketch of the companies that you've worked for, your experience in IoT. 

Silviu Neghina: Yes, thanks. I started working 20 years ago, mainly around, you know in the telecom industry for Orange, and then I moved to my preferred topic to the IoT, and I started working 13 years ago in the IoT industry for Orange. Then moved to Sigfox, then to LinkSense. And nowadays I'm working for Vizio Sense. 

Nick Earle: I was going to talk about Link Sense cause they're in a smart label field and we'll also talk quite a bit about your current company, VizioSense. Actually, before I do, maybe I can just ask you, Sigfox is an interesting company. Certainly, a lot of our European listeners will know quite a bit about Sigfox. 

What's your take on. I mean, it was a fantastic idea. What's your take on the advantages of Sigfox and where they are as a company? It's a collection of companies really, but what's your take on Sigfox before we get into the other stuff?  

Silviu Neghina: Well, for me, Sigfox was the new kid on the block, and I think it was the technology that started to ramp up the IoT with real numbers. So, I'm excluding now all the analyst estimations, and all the billions of devices that were supposed to be active. So, I think Sigfox as a technology helped a lot for the whole IoT industry grow. I don’t know exactly where the company is today since it was acquired. Of course, I'm still somehow emotionally attached to Sigfox but overall, I'm emotionally attached to all the IoT industry. 

Regarding Sigfox, I think it started with a huge ambition. It didn't turn out that great in terms of reality versus all the forecasts, just because I think it's a little bit complicated to align more than 86 Fox operators that are managing every operation in a country and make them all think the same, have the same strategy, and scale the technology at the highest level. 

I think that Sigfox based in the insolvency issues in our era, very close to the pandemic. And I think it, it was not the right moment for the IoT industry and for everyone at that stage. But overall, I think it's a great technology. It's a great company. I hope the technology will scale even further and all the Sigfox operators will keep doing a great job because I think the IoT industry does not need technologies that are disappearing. 

It needs technologies that are staying alive and scaling. And reaching the billions of devices that we all dream about.  

Nick Earle: There’s that potential that we talk a lot about on the IoT Leaders. And we're going to get into that. And by the way, it's remiss of me because I do know something else about you, which I didn't mention during the intro. 

I think you're also our first guest. We've had over 50 getting on for 60 guests on IoT Leaders. I think you are our first guest who could lay claim to at one point, who made their living from playing tennis. Is that correct?  

Silviu Neghina: Yes, yes. It's my first love. I've been playing professional tennis at junior level, right when I was a kid in Romania. 

Unfortunately, back then in the nineties, I needed to choose between school and sport. It was an obvious choice, but I love playing tennis. I play, even nowadays, I play tennis in a semi-pro tournament with people around my age. I'm a very competitive person, both in the personal life and in the professional life, and it's a sport that I really love. 

One of my idols is Roger Federer from the tennis world.  

So yes, I played a lot of tennis when I was young, and I encourage everyone that has kids to push them to professional sport no matter what the sport is, because I think it only brings benefits for everyone. So  

Nick Earle: I'm sure we could talk a lot about tennis, and I have had the privilege of seeing Federer play live a few times. I always say he didn't have a racket, he had a wand. 

It was just magic what he could do. But anyway, let's get back to IoT. So, let's start off. Why don't we start off on LinkSense? Let's do it in chronological order. I mentioned earlier LinkSense in the smart label business. What's your take on the smart label? It is a hot subject. People have been talking about massive IoT. 

You mentioned that the forecasts that were put up, the billions you said that were missed. I mean, the famous quote, I've referred to it many times is that there was going to be 20 billion things connected and there were only 11 in the timescale that we said 50 billion things and there was only 11B. 

So, classic Pareto, but actually when it came to smart labels, you're talking massive IoT, you know, 500 billion things, a trillion things. The ability to track everything at a much lower level. And so, what's your take on where Smart Label is right now and where it will be in the future?  

Silviu Neghina: Well, I think the smart labels nowadays are at the same stage, like the whole IoT industry. 

People are focusing a lot on the technology that is being used, and there are many players that are not fighting, but anyway, claiming somehow that the technology that they are using is the best. When we talk about smart labels, for me, I think there are two major categories is the one that has really scaled up and we talk about billions of labels per year. 

The ones that are connected by RFID and or NFC, depending on what the use case we are talking about, but mostly about RFID. Here. There are some major names that are public already, so I'm not disclosing any sensitive information like the cat loan that is connecting all their assets in all their stores. 

In Romania, we have all the stores where you just purchase the products. You just put them in a small box and then the invoice is automatically generated, and you just pay with a card and that's it.  

Nick Earle: So, let's, let's just double click on that. That's really interesting. So are you saying that you go into the Decathlon, the big retailers say sporting goods or whatever, you go in, you buy a pair of trainers or something, whatever. 

You put it in a bag, and you leave the store?  

Silviu Neghina: No, I go to the self-checkout. I just put them in a kind of a box that has some RFID antennas inside. The products are automatically scanned and then I just click on pay with a contactless solution and then just leave. 

So, this is the technology that scaled a lot. The smart labels, there are different people that are not really calling this label smart labels, but for me they are smart labels. The moment when you are almost autonomous in a store, that's definitely a smart label. Now there is also the other area when we talk about active smart labels, and here we talk about Bluetooth, low energy. 

We talk about Sigfox, we talk about LoRaWAN labels, and we talk about, let's say, the premium smart labels that are connected by cellular. For me, I think first of all, in general in IoT and also in smart labels, I think there is no technology that fits all the needs. We as solution providers, the system integrators, telecom operators, everyone in this industry should stop selling technology. 

The customers don't really care how they get the data as long as it's at the lowest cost and it provides them the fastest return of investment. So, all these stupid fights in the industry about the asset: “Sigfox is better than LoRa. No, LoRa is better than cellular. No cellular is better.”  

Actually, no technology that is better than the other one. Only about the availability of a network because you have certain countries where you have Sigfox covered completely. You have countries where LoRaWAN has covered most of the country, or at least the urban area. Of course, you have the telecom operators that did good jobs so far in the industry. 

My take is that they could have done a much better job, but that's a separate topic and you have all these technologies. I think it's about the type of good that you want to connect, the return of investment calculator that you provide to your customer and your partner, and then, and only at that moment you choose the technology. 

Because when we talk about smart labels, we talk about, of course, non-lithium batteries. You cannot put a label that is disposable and works for one week or two weeks and put a lithium battery. I mean, we all talk about saving the planet and reducing the seal of footprint. You definitely cannot put lithium there. 

Of course, you end up in a situation where everybody wants to save the planet until they see the cost of it, because of course, a non-lithium battery, a printed one, it's much more expensive than a classic coin cell battery. That's why I believe that if we do a consultative sale and not the transactional one. 

Then we pick the right solution for the customers, and we are really able to scale the labels. And going back a little bit, one step backwards to your initial question about the labels. I think nowadays it's a matter of who is really able to produce massively a smart label at a very low cost. Here, I think it's still room for improvement from different suppliers. 

Nick Earle: And can I jump in? We've talked a lot about the inflection point of adoption of all technologies. IoT just a collection of them. As you point out. It's not a single thing but there's always, in every, for the last 40 plus years, there's been an inflection point of adoption of every product, every product, and prices is. 

Absolutely a key to the inflection point. What's your take on putting aside the different technologies for different use cases? The point that you're making, but in terms of a smart label on a box or something that moves, you want to track it. because you want to look at the supply chain, especially with food, I mean, you know, you've only got a certain amount of time before you have to throw it away. 

What? What is that price point? What is the holy grail of the cost of the label, if you like, that will really encourage massive adoption?  

Silviu Neghina: Well, I think the massive adoption of active smart labels will be in mid to long term, but the price point of one Euro, I know it's very low and I know all the industry will hate me for what I'm saying now, but I think one Euro will be the price that will help us scale. 

And be able somehow to match the RFID because you have an RFID label that in use cases like decathlon or the big ones in billions, I think it's even less than 1 cent per label, which is almost nothing. I mean, it's like the numbers after the comma in Excel.  

Nick Earle: Let's call it a hundred times below the target price. 

Silviu Neghina: Yes, but on the other hand, you get much more relevant information from an active label that for most of the use cases, it'll make sense to pay that one Euro. Now, depending on the technology, we are pretty far away from that one Euro. I think where we will be able to design a device that is less complex in terms of architecture that is using a very efficient module, it's using a very efficient battery and the R&D teams, because I think the R&D teams are critical here, and if you have a competent R&D team, you can try to reach those low points in pricing. 

So, you’ve got to anticipate very well because if a customer says that they want an autonomy of three days, one day above that is increasing the cost of developing of development for you. So, I think we will scale massively with active labels when we reach one euro give or take price range in the short to midterm. 

Nick Earle: I think an inflection point is around 10 to 15 Euros. Right? But then we are able to connect the specialist use cases for high value goods.  

Silviu Neghina: Yes. I think definitely pharma and medical. I'm not saying that those are industries where the price is not important because of course it's important. Every industry is important, but when you ship a bunch of drugs for cancer, for example, that are in place in a small box, but the price of them is about a million euros. 

Then I'm sure that 25 per Euro, that is the price today of a cellular label shouldn't have a major impact once you have a full visibility on the whole supply chain. Yeah, so I think it's important again, and I'm getting back to the consultative sales. Let's pick the right device for the right use case, for the right coverage, and only then choose the technology. 

The technology is just a resource, it’s a tool. It is not something that should be put in front of the customer because. Let me give you an example. I've attended a great meeting organized by a partner with some of the biggest names in that industry. I don't want to name the industry because it might be too relevant, but I realized that we attended that meeting trying to present the moon and the sun and the stars, and actually the guys weren't even using a basic RFID label. 

They did. They weren't understanding exactly what means to connect, but the device as an asset. That is low cost that today is being scanned or being written in Excel, the ID, and we were pitching smart labels and short detection movement, temperature and all of that. I think we’ve got to do a much better job in identifying the customer needs and then build a solution, work with the customer, provide the good price so that it's going to be a win-win for everyone. 

Because if you go with huge margins just for the sake of improving your own P&L, nobody is going to sign for it.  

Nick Earle: IoT is what we're talking about now, but we could go back 40 years and just talk about sales, minicomputers, PCs, LA laptops, phones, software licenses. 

It is a consultative cell. And you mentioned the ROI and really getting to the ROI and then mapping onto the technology. Some people would say that in IoT you have hundreds and hundreds of partners. Yeah, but a lot of them are just pushing the technology. They're “this is what I have to sell, and this is what you need.” 

So how much of a challenge is it to get to the ROI on the use case? We'll probably talk about that in terms of your current company, VizioSense, when we transition to that. But in terms of smart labels, how difficult is it and what's your experience of where the ROI is? Because that will determine what you can afford and that will then determine what technology that you can use right now. 

Silviu Neghina: Well, I don't want to deliver some theoretical information because I'm always against this, this approach where we just sell dreams. I'm going to give you a real example from a former discussion that I had. So, in the past, I was talking with a major brewer. I think we all like the beer that they are providing. 

And we were talking about the percentage of loss in the refrigerators. And it was a major issue for them. They were losing some double-digit percentage of their refrigerators year on year. Of course, the first question is, okay, but we are living in 2022, I think it was the year. So, we started to do some steps in connecting assets and all of that. 

So why this issue? And the main issue was that everybody was looking just at the initial cost. We have a million assets price for five years is, I don’t know, 50 euros. So, we got to invest 50 million euros. It's huge. No one will approve it. Of course, if you put it like that 50 million euros and you go to the board and say, “guys, we need 50 million euros to connect some assets that hopefully will improve our business.” 

Everyone will say no. But moving a little bit to the return of investment, I was explaining to them that once you connect an asset like a refrigerator. You gain some benefits, like first of all, you see the real time location of it. In theory, no asset should move unless you give the approval to move it. 

Secondly, you get the information about the temperature. Okay? For beer, it might be not that critical. I mean, we all love beer, but if a beer goes from cold to hot and then back cold, I might not be able to distinguish it. But if we have ice cream, it's a totally different story. 

Then so you monitor location, you monitor temperature. Then if you monitor with the same device, the number of times the door is opened. Mainly when we talk about transparent doors, then you could map this information with sales in that specific store. And then C, maybe the store has put other products in your refrigerator, which is forbidden because that's why you are giving it for free. 

Maybe there is an issue with your offering or with your merchandising. I don’t know, maybe your labels with the pricing are not visible enough so that people need to open the refrigerator. Maybe there is an issue with your pricing. Maybe you are overpricing your beer versus your competitor. So, I see all of this as advantages. 

Now when you calculate the return of investment, of course you got to look at the expenses that you have with the people that you need to send on the field and okay, maybe countries, like Romania, Croatia, Serbia, Czech Republic. Maybe we talk about small countries, but look at Spain, look at France, look at Germany. 

These are very big countries. In order to send someone to check every refrigerator from time to time, like a merchandiser or any other road. You have huge expenses and it's not, it is not just about the person, the salary and the benefits and all of that. You’ve got to provide a car. You have expenses associated that. 

So, it's a huge expense. So, once you calculate all of these benefits and you see what you can save from this, then maybe those tens of millions of euros, that will be the cost for all the five years. Maybe it won't be that much. You need to look at it from a very constructive perspective and calculate to the customer this return of investment. 

And then you will see that it's not an easy decision because no sales are easy nowadays, mainly in IoT. We talk about a long-life cycle, but still the customer will understand why he needs to pay something, and we will understand the real benefits, and you will put there a figure in my activity. You are asking about labels at LinkSense. And also, nowadays at VizioSense, I have built return of investment calculators for the customer. Of course, you got to do more a deep dive into that specific context, but I'm trying to show them from the first meetings based on some basic information that I get from them guys, look, this would be an approximate return of investment for you if you decide to work with us. 

So, I think that's where the difference should come from as a solution provider, or system integrator: that's the value.  

Nick Earle: And we agree, and I'm going to now transition to VizioSense because it's a very interesting company, the company that you're now at. 

In fact, I think about 10% of people watch the YouTube a version of this. We have coincidentally released an Eseye whitepaper on how to get to the ROI of IoT connectivity for exactly the same reasons as you said, because a lot of people who are buying connectivity are saying, well, the cost per megabyte is this, or whatever. 

And of course, the data with data prices collapsing for IoT by about 15% a year and have done for about six years. If you Google gigabytes of IoT data over the last six years, you've seen an almost perfect straight line. And you know, the idea of saying it's all about the cost of data, which a lot of requests come down to. 

“How much is your data; how much is your data?” When actually data is a tiny, tiny percentage of the cost of the project. And I'll just use one quick example before I get onto VizioSense. In our case, I'll share an example with you. You know, we do, for instance, all the EV chargers for Shell. 

Shell Recharge is the name of the company. You see them all around the place, you've got the AC chargers, but you've also got the DC charger, the Rapid Chargers. We make a hardware router that's in those DC chargers with the connectivity, but the thing is that they can make that unit. In one factory in Vietnam as a single product SKU and ship it around the world, and it just connects well. 

The cost of not having to have more than one factory or more than one manufacturing line in one factory is 10,000 times greater than the cost they're spending on the data, and yet very few opportunities come to. As an IoT company with hundreds of customers, very opportunities come to us with an ROI built into the initial request. 

It normally comes from the product side of saying “how much your SIMs and how much are your data?”. So, I think as an industry we definitely need to go through that maturing side. And I know that you are doing that right now because that's a great transition. In the area of smart cities and video cameras in VizioSense, you've taken that philosophy forward into your current role. 

So, let's bridge now. So why don't you tell everybody a little bit about VizioSense, what they do and what you see going on in that space.  

Silviu Neghina: Oh, I joined VizioSense at the beginning of this year. I'm working with a former colleague from Sigfox called Maxim, who's the CEO of the company, and I discovered a great solution for smart cities. 

Mainly around the smart parking and urban mobility. And on top of that, some what I call mini spinoffs of analytics for retailers, shopping centres, logistics, and warehouses. So, we are a French startup that is manufacturing and designing in Europe, which is a major differentiator nowadays.  

Nick Earle: Yes, that's a very topical subject nowadays. 

Silviu Neghina: Yes. But we see many news stories about different governments that are forbidding different Chinese manufacturers in their networks. I don't want to go into the political discussion, of course, but we are designing and manufacturing in Europe on an edge AI sensor with one or two cameras. 

We are analysing parking spaces, monitoring traffic, and counting people, objects and vehicles in very dense areas, which is a major challenge. I mean, imagine a concert with a hundred thousand people and having an accuracy of more than 98%. It's something absolutely impressive and we have an AI team that is doing an amazing job on a daily basis. 

And now all the pressure is on us, the commercial teams to really scale. But why I think it's that impressive in what we are doing. And it's not just the fact that we are Vizio, but it's important that we are working in scenarios where the return of investment is very easy to calculate for a municipality. 

So, we are all, almost all of us probably are drivers. We are driving more or less depending on the country where we are living. For me, in Romania, for example, I'm driving quite a lot. In other countries where there is a much more developed public transfer problem and probably it's a little bit less, but still, you might need to drive. 

Now imagine how good it would be when you go to a meeting to check out all the parking spots around that area to see if there is any available parking place. Check some historical data and see, okay, tomorrow middle of the day in the centre of London. Historically speaking, were there any available parking place at 11 in the morning? 

Yes or no? And you see the chances of finding one park parking place and also seeing the real time availability of them. Now, when you put all of this in perspective, you get less traffic because in a normal world, if you see that there is no parking place available. And historically there were very, very low chassis of getting a parking place. 

Probably you don't need to drive. That place. So, in terms of return of investment, you have less traffic. You have a huge customer, a much, much better customer experience for your citizens, or you as a driver, you go there much more relaxed. You don't stress yourself with finding a parking place, being late to a meeting. 

Now, from cost perspective, what we provide versus the cost of paying one hour per spot. In, in any city in Europe, there are not differences that big. The return of investment for the whole solution for five years is in less than one month. And I challenge anyone in Europe or even abroad to do this exercise. 

So, when you have a cost of few Euros per hour per day, and an estimation of few hours of parking space is occupied, like, I don’t know, eight, nine hours. Then you look at the cost where parking space to monitor it with an NJI sensor, so here we don't talk about classic parking sensors where you drill a hole and then you put a sensor and then you calibrate it, and then you need to make sure that you have a good coverage in the area. 

And then if you do some civil work, you’ve got to unplug it, then re-plug it, reconfigure it, and all of that. That brings a huge cost. With Vizio’s solution, you just have a sensor that is scanning and on average it gets 20 to 30 parking, place parking places. In a city, you have a cost that is one digit and a very low one per month, and that cost is being covered by the first few hours. 

Nick Earle: So sorry to cut across you. Let me make sure that I get it. And also, the listeners get it, is the ROI and you mentioned one month. That's very impressive and is that because you get a higher utilization of the parking places, if you are the municipality that you are optimizing people are, people will go to those spaces because they know they're likely to be available and therefore the overall occupancy rate across all of your spaces rises up and you get revenue from that. 

Is that the basis of the ROI?  

Silviu Neghina: That's the basis of the ROI. And then this is an action that takes place in the first six months mainly, and then you reach a default mode. When people are looking for a parking space, they see the availability. They start being used to pay for it because most of the municipalities nowadays, they have an issue with the cars that are parking. 

They pay the first hour and then no one is paying the difference. That's in the ideal scenario when they pay the first stop.   

Nick Earle: Excuse me, they can't afford the wardens to walk down the street, checking and making notes in a notebook. And yeah, the use case goes out the window. So, people are aiming the system by saying, oh, nobody will check. 

I'll pay for it. I'll pay for one, but I'll take up three.  

Silviu Neghina: Yeah, it's a matter of course, it's a matter of luck in the end because you might have a policeman that is driving by. 

Nick Earle: I was thinking of my apartment in London. I don't think I'd gamble on 10 minutes given how keen they are. But that's a particular part of London. 

But in general, yeah, absolutely. People will. 

Silviu Neghina: You occupy in a more efficient way, the parking spaces. At the same time, make sure that your citizens are paying for it. And then of course you can start playing with dynamic pricing, with subscriptions, with anything you want. What as a basis for the first month, or let's say the first year, you make sure that every parking space is generating revenue, and then of course we move to the parking spaces for electric vehicles. For disabled people.  

And then you make sure that they are being used by the right people. So, it's digitalizing an asset because the parking space is also an asset at the end of the day. This brings a lot of revenue for the customer. If they are doing the right thing and if they are willing to listen, and if you as a supplier are doing a very transparent job for them in showing them everything from expense to maintenance costs because those are also important revenue that they can generate from that parking space. 

I even had discussions with mayors that say, “Ah. Wow, it's that fast, the return of investment is super.” So, in that case, I can even decrease the cost because I still make enough money so I can decrease the cost of the parking. People will like me more and I will still make a lot of money for the municipality, and then I can reinvest them in schools, hospitals, public parks and everything. 

Yes, it's that easy. Of course, public sector is never easy, so. I don't want your listeners to believe that. Yeah, we go tomorrow and in one month we sell. No, it takes a lot of time, a lot of work. But that's the beauty of IoT I think at the end of the day. And that's coming from an IoT addict. You work, but you have a huge value for everyone involved in this picture. 

Nick Earle: Alright, so you need a camera? You mentioned that you need a sensor. So where does the camera go? Who puts it up? I mean, does it need to go up a tree or how does it work?  

Silviu Neghina: We provide the full solution, the imagine the camera associated with your own eyes. 

Whatever the camera sees is whatever that information that the device is processing. So, we put it on streetlights, we put it on buildings, we put it on billboards, we put it basically in every possible place above four to five meters, let's say. We start scanning the area based on the products that we define that need to work in that specific area, and then it simply works lifetime up to 10 years. 

So, no issue about that.  

Nick Earle: And what role does AI play in this solution?  

Silviu Neghina: The AI plays a major role and that's why I'm so happy for being able to be at the right time available for VizioSense and to be part of the team because I feel mixing AI with IoT is something that is the future. Probably for our industry is, I don't want to say it's the game changer, but it's the future because it combines two great topics. 

So, the AI comes in place when we analyse the data. And we analyse it in the device. So, we don't store any video, we don't store any picture, we don't send any data to any cloud. We analyse it and then provide the information. Everything is done with the lowest possible cost because we send few kilobytes. 

The only megabytes that we are using is when we are doing some software updates, but besides that the traffic is very low, even though we are using a classic 4G connection. So here we are not even talking about low power wide area networks. No, a classic 4G connection. And we analyse the data with our AI models in the device and then to the customer. 

We just send the relevant information occupied or available, the increment when we count people or vehicles or the number of people in major gatherings. And also, I was saying earlier about the spinoffs and the way we use, we will, we are constantly challenged by our partners and our customers. We've developed new products based on our core competencies, and nowadays we are able to replicate what we are doing for a car. 

We do it at the same time for a warehouse, and we monitor the time spent by a truck to load the goods and to live in this way. It's a major improvement in the logistical process for warehouses and logistic companies, and we are doing the same type of information. For retails and for shopping centres with our sensors, you get the full flow of the people in a very privacy-oriented way. 

Because we are fully GDPR compliant, we don't see faces, we don't deliver gender information, nothing like that. We just do the basic information, but the most relevant one in the most private way possible because privacy nowadays, it's a topic and it becomes a more and more discussed topic by European Union and by everyone. 

We want to make sure that the products that we are delivering are respecting the people's privacy and are delivering a huge value to them. And so far, I think we did a great job.  

Nick Earle: I was just wondering, listening to you, I'm not familiar with the situation in Bucharest, but if I compare it to a lot of smaller towns, let's say around the UK. 

They're very concerned about the death of the high street with Amazon, and I don’t know how advanced Amazon is in, well, you operate all over Europe, so you just happen to be based in Romania, but you fly to work. Because your office is in Paris. That's something I've flown to work for about half my career, but in general there is, I'm wondering if there's a link not just the revenue generation for the municipality, but also to societal link in terms of protecting the high street. Because what's happening, certainly over here where Amazon is, you know, there are Amazon Prime vans everywhere and Amazon's a customer - full disclaimer. Amazon's a customer of ours and we do the EV charging for the prime vans and we do the automatic door opening for Amazon Key for Business. which I've covered on a previous podcast. 

But actually, the more that Amazon grows, the less people obviously go to the high street. And it's not just that, it's the fact that it's such a pain in the butt going to the high street, certainly in a small, crowded place. It's not like the shopping mall that I was recently on holiday in the US you had massive shopping mall with huge car parks where you can have 5,000 cars. 

It's not like that on some of these high streets. And, and if parking is a pain, you think the next time I'll just click and have it delivered on Amazon Prime. So, is there, do you find that's a factor yet or not really? Is it primarily around the revenue generation / redistribution of the money to other initiatives or are people saying to you, this is an important thing to do for society, to preserve the nature of the community around where the shops are, where the parking is needed? 

Silviu Neghina: Well, in our ideal world, I'm getting back to my quote about labels. When I said that everybody wants to save the planet. There are many, many people that are concerned about the non-revenue things, but at the end of the day, it's all about money. Somehow that is generating, that is saved that money. 

Yeah. I think the whole experience is also important for customers because take for example, the airport parking, the long stay parking. In theory, if you would talk to them about VizioSense or other players like us, because we are not the only ones. But if you are saying, guys, let's try to digitalize what you're doing there, they will say, “No, we already have customers. People are already paying 17 pounds per day. It's a pretty good business. Why would we invest?” 

And the answer is pretty simple because you have a parking space with 5,000 spaces and only 20 of them are available. When your customer is paying 17 pounds per day, I think he would be more than happy to see where he needs exactly to go so that he would make sure that they are not losing their flight because they are circling around just to find that specific parking space. 

So, at the end of the day. It is about money, but it's more about the customer experience, the experience that you provide to a citizen. So, there is a strong link between that and the experience. Of course, there are some that just look at the dollar sign and that's it. 

Nick Earle: Yeah. Okay. Very good. So, VizioSense, smart cities. So, you mentioned car parks. You've mentioned parking. Maybe one last question. In terms of the sort of things that you're seeing, is there an example of an unusual use case that you've seen where you, somebody said, well, could you do this? 

And you hadn't thought about it, can you?  

Silviu Neghina: There are many examples, and I think the strange examples are coming mostly from the false idea that the AI can solve everything. So, I would say the craziest talk that I had about these solutions is being able to count every material on a construction site. 

And I said, okay, but can we define the number of items that you are using? Yeah, we have about 20,000 items that are being stored in most of this area. And they are one like that, one of that type, one of that type. And they're all thrown there, and can you count all of them? And I was like, man, okay. 

I would love to do that. But unfortunately, we are far away. But I like to be challenged and I'm the one that is challenging my AI team a lot, but in the use cases that are close to reality. So not the ones that are impossible to solve. But in one very interesting use case and the link to this green approach that we have at least in mainly in Europe, but probably in other places also. 

We are developing now a new AI model for a partner where we look at the wheel loaders that are being used in recycling plants. We are helping them count and estimate the type of recyclable garbage that it's in that specific plant and the activity so they can improve their processes and generate more revenue from this operational flow by making sure that the operators from every specific waste plant are doing a better job and are the right flows in the plant. 

So, there can be efficient. So now we are developing a new AI model that is not only counting a wheel loader entering or exiting a premise, but it's also looking if the front of the wheel loader is up or down, meaning if it has garbage or not, and how many times it crossed. We anticipate that the number of tons, because you could estimate it based on the papers that you have, based on the visual that you see. 

But it's not a precise estimation there, you know, and this is the type of use case that I like to do because it's closer to reality. Of course. It takes time to be developed. Also, another one you mentioned, you mentioned some names, from your current experience. Some of them are also in our prospect list or in our pipe, let's say. 

It's great to be able to also from logistical per perspective, not only to check the time spent by a truck when loading but also imagine putting our sensor on the other side inside the warehouse and count every returnable transport that is being loaded in the truck. So, I'm an operator. I know that I got a load 17 pallets in a truck and I load the first one, the second one, the third one. When I load the 17 one, I get a green light and then I can say to the driver, okay, you are good to go. This doesn't mean that with our solution, we know exactly what goods are there because the goods are connected by the smart label that we are talking about earlier to know exactly if that precise good is on the right pallet. 

But in terms of operational processes in a warehouse, it's very important that the guy is moving very fast and knows exactly what he's been doing, because if the truck is leaving and it has minus one pallet, then the costs are huge to get it back loaded again and send it back on the field. So, these are the types of use cases that are not off the shelf. 

Our AI model is not trained to do them, but my team can train it very fast to detect that. But we’ve got to be very realistic. The AI is not solving everything in the world. So, some of the people talk to the ChatGPT nowadays, like it's their best friend. Sometimes it provides good enough information. 

But when we look from B2B perspective, the AI is very good, but it needs to be somehow trained into a specific direction.  

Nick Earle: So, a closing comment from me. It is interesting. I've been doing this podcast for four years and I would say for three years we didn't talk about AI. Maybe a little bit more than three years, even though the industry was talking about AI. 

But nobody ever came on to the pod saying, well, we're using AI and in the last six months that's totally changed. We've had some people who have then said, look, it sounds obvious, but AI's only as good as the data you feed it. And the question is, where'd you get the data from? It isn't just scraping the company's records to find your internal documents. 

It's the data from things that you would need. And we had a guy from Volvo on, I guess two weeks ago talking about Volvo's 150 factories and you know Millions and millions of things. They want to connect basically bits of machinery to keep the production lines flowing and to never do proactive pre-emptive maintenance. 

So, the manufacturing lines never stop. But the point being is that that is a problem that. It can only be solved by collecting data from everything. And to your point, using different protocols, different technologies to get the data, because you're talking about lots of different things that were made by lots of different companies and there isn't one solution that fits all. 

So, it's always going to be a patchwork quilt of different technologies. But what we are now seeing, and I think in general; we are now seeing a growing realization. That there's another type of benefit for AI, and you've mentioned it and the guy from Volvo two weeks ago mentioned it. There's another type of AI, which is AI in the industrial sense, the AI with data from things which hasn't had anywhere near the amount of publicity that that AI has had in other areas. 

You know, lawyers are going to lose their jobs because AI can read the contracts better. We're going to be able to do healthcare better. There's going to be mass unemployment for graduates. When they leave, because the jobs they used to do in the first two, three years are all now done by AI. But now we're starting to see, it's only just, I would say it's, you could argue it's always been there, but I think it's starting to get much more airtime if you like people saying. 

The real opportunity for AI. Because there are many, many, many, many, many more things than there are people in the world. The real benefit of AI will be getting the IoT data into AI as opposed to the question we were asking nine months ago is what's the use case for AI in IoT? Well, actually it's how is AI improved by getting a thing, data about things into AI, and then there's a whole series of benefits which are very tangible. It comes back to your ROI point. It's really hard to be tangible about the return on AI in an office environment. You know, its efficiency, but most of the people are still there. And in the university environment, well there is no benefit. 

They just use it to cheat on their exams is how the universities think about it. But in the industrial side there is linking to the ROI, if you have an ROI based approach and you can actually do things more efficiently, whether it's the proactive maintenance or pre-emptive firmware updates as you said, or optimization of car, car park spaces for drivers, or making sure you get most revenue from parking spaces on the street. 

It seems like there is a whole new awareness of an even bigger benefit of AI when you think about the billions of things, and back to your point about smart labels. Potentially trillions of things in the future that you'll be able to get data from. And if you can get data from trillions of things, and with the AI models getting better at between five and 10 x per year, which is an incredible number, then suddenly there's this whole new set of benefits and opportunities and use cases, probably companies that we haven't even thought about because we were never able to connect. 

Those two things together, we were never able to connect. We were just connecting and collecting basic telemetry data from things, if you like, or usage data. But now with all the things we can connect, it does seem like each guest we have here on the podcast is now saying, I am.  

The AI models that we are developing for our customers and that we're seeing a lot of that right now and it's, I don't think it's made the mainstream media right now, but I think it will do because there's a lot of people working on some very exciting projects around feeding the data from things into AI models, and that's going to be a very exciting few years of which you're one of the companies doing just that. 

Silviu Neghina: Yeah. And I also saw a trend in the last weeks or months of people praising the death of IoT because now we have AI. For me that sounds really stupid. I mean, if you say something like that, it means you really don't understand anything. I get it. And I remember you were asking me, the first question was about Sigfox, how it was, I remember in theory a competitor of Sigfox that was very happy that the company will die. And I was trying to explain to him, man, it's not important that you are building LoRaWAN base stations and a competitor. Technology might die because it's an issue for the whole industry. If tomorrow Sigfox disappears, you will sell fewer base stations because people will lose the trust in IoT. 

Also, same stupid approach nowadays with AI will kill IoT and AI will still be there and the IoT is dying. No guys, AI without IoT is useless. That's what I said. So, the IoT is the one that is generating all the data to train the AI.  

We feed the all the AI models, at least in the B2B and industry scenarios. We feed them with data coming from connected devices. Those connected devices are the basis of the IoT nowadays. So, let's not try to shift this also into a technology fight, and let's accept and understand that AI plus IoT is the future. 

The IoT was the start. It was the game changer for sure. Some of us did a good job. Some of us, I don't want to say a bad job because I don't think anyone has done a bad job in IoT. Some could have done a better one, let's say. At the end of the day, these whole industries will reach their highest potentials. 

By combining the AI and the IoT in few years, we will not think about should I connect my beer keg to get it back faster and sell more beer? No, it's going to be what kind of decision my new AI model will take so I can improve. You will start applying some numbers and say, okay, I have an objective of two x revenue increase in the next three years. 

What do I need to do? And then you will train your AI model based on your specific business to take some decisions for you and your teams. So, AI plus IoT is here to stay, and I are here to change our lives both in private and professional way.  

Nick Earle: Absolutely, and that's a great place to finish, Silviu. The purpose of IoT is going to be to get the data, your data, not public data, your data that you can use to train your IoT model, your AI model that will optimize your business and make it more efficient. And that is why IoT has to be very high up on people's priority lists because it will get you the data to train the AI model in a way that generic AI models won't help you because they're not being trained on your data other than some documents that you have that may need based on publicly available data. 

So, I think that's going to be a theme of the podcast for some time to come. And it is interesting how it's changed in the last six to eight months in this space. Listen, we could talk for a long time. But we are at the end. I want just to thank you for joining. I know you've reached out, said you'd love to be on the pod. 

I really appreciate that. And if people want to contact you, what do they do? Do they go to VizioSense? What do they do?  

Silviu Neghina: I think VizioSense.com in terms of VizioSense products and services. My LinkedIn profile is active, and up to date. I'm a very social oriented person and I answer very fast to any kind of request. 

Like I'm saying to everyone, there is no dumb question. There is that question that hasn't been addressed, so challenge me with any topic you want guys, about IoT. I'm an IoT addict and I'm happy to help.  

Nick Earle: Okay, and it's Silviu Neghina, N-E-G-H-I-N-A at VizioSense. Silviu, thank you very much for being on the podcast. 

Good to talk to you. Good luck with the commute to and from Paris and all around the place. And hope you still find time to keep playing the tennis at that high level.  

Silviu Neghina: Thank you very much for having me here, and cheers to everyone. All right?  

Nick Earle: Yes. Good to talk to you. Thank you. 

Outro:
You’ve been listening to IoT Leaders, featuring top digitization leadership on the frontlines of IoT. We hope today’s episode has sparked new ideas and inspired your IoT and digital transformation plans. Thanks for listening. Until next time!