Mister Beacon Episode #6
Nextome CEO InterviewJuly 03, 2016
Vincenzo Dentamaro reveals details of their deployments at Fiat's factories, the Rome subway and museums in Italy and the US. They talk about their approach to pricing and their future plans.
Receive new Mr. Beacon episodes right in your inbox
You're listening to The Hitchhiker's Guide to the Beacosystem with Steve Statler.
Vincenzo Dentamaro 00:33
We are the first company mapping over 1 million square meters that are old row medical stations. We have a lot of museums and all these things are public. So the idea was back in 2012, how we can know let's grow the cultural heritage of our place. We mentioned that technology there was capable to filter out noise coming from because there are a lot of robots working. So iron everywhere. No, thanks very difficult for our filtering engine.
Steve Statler 01:24
Hello, and welcome to episode six of The Hitchhiker's Guide to the Beacosystem. My name is Steve Statler of Statler Consulting and we're doing this in partnership with our friends at Proxbook. And this week, it's off to Southern Italy, talking to the founders of next to me, Vincenzo Dentamaro, who's the CEO talking about all sorts of deployments, they've been doing it Fiats at the Rome subway museums. And we're gonna get some really interesting insights into their technology into their business, they'll talk pricing, and some of their future plans are going to be going for a round of funding soon. So have a listen in hope you enjoyed as much as I did. Maybe we should start off with some introductions and you tell me who you are and what you do?
Vincenzo Dentamaro 02:11
Sure. I'm a software engineer. Also, I founded with Domenico, j&j, and an angel investor called the market people, typically the startup in the background doesn't 15. But our research on in that position is started in 2012. Since three years now, we started to work on that company, because we think that the opposition will be the next big thing. But I think that it would be something like that, all right, like, you know, search engines back into in 1997. In darkness, you position will be the same, because, you know, we have a lot of buildings. And when you get to indoor, you're, you're blind completely because there is anything that works. So we have invented this technology to solve this kind of problem. You guys are based in Italy, I think I was looking at Proxbook and there's we've got a dozen companies in the ecosystem who are working in Italy. I live in America, I grew up in England. And so our view of Italy is very narrow. We kind of think of Rome and Venice and maybe Milan, and that's about it, but you don't live in any of those places. Where do you live in work? We are in the south in the southeast, the Pulgia. Alright, so nearby bodies, the city on the sea that is in front of Greece, so we can get on grease on other kinds of places.
Steve Statler 03:44
And so are there many technology companies in your city?
Vincenzo Dentamaro 03:48
No, not so much. We are in the south of Italy and south of Italy has to grow in the sense about startups ecosystem, and so on Italy's early startup ecosystem, it's not the valley, let's say. We were in Silicon Valley, but Italy is clearly but we also have a lot of talented people here to bring on board.
Steve Statler 04:13
Very cool. Thanks so much. And just to kind of orientate everybody you do indoor mapping, is that kind of a very?
Vincenzo Dentamaro 04:25
Well, we do indoor positioning and navigation. Okay. Yeah. So we mean, it means that we can find the user's position indoors with one 1.5 meter of accuracy.
Steve Statler 04:37
Okay. That's very accurate.
Vincenzo Dentamaro 04:40
Yeah. And now if you want I can explain your why. Well, we start with anytime we have a lot of museums, and all museums are public. So the idea was back in 2012, how we can you know, let grow The cultural heritage of our place. And we invented the with an app that was capable to know user's position indoors, and show advanced contents based on position. So let's say if you're in front of Mona Lisa, and you want to get the description or a video, introduction of that you don't have to, it was really easy. Just get very close to the Mona Lisa. And you get to be there. That seems to be proximity. But it was, because if you put becomes very close one to each other, what happens is that the noise coming from from become probably you will get opening a video or a content of another piece of artwork, because they are very noisy. And we started working on that. We called in our patents that now is a PCT International Patient patent. Also, states in Europe. Singapore, we, we invented a technology there was capable to filter out noise coming from beacons, we call because before those BLE sensors, and basing on that filtered signal, we can get an accurate position of the user. And you guys have videos on on the web, and I've looked at them, and it really does seem like you achieve very good fidelity, how many beacons do I need to use to get that kind of accuracy that won't help meters accuracy? It's clear that the accuracy depends on the number of beacons. But we have estimated that to achieve one 1.5 meter of accuracy, it's necessary about one beacons every 35 square meters. But if you want normal accuracy that is below the three meters, also, while we can every 50 square meters is okay. Okay. And so do you have many customers? We do have big customers here in Italy, initially, and also one in San Francisco. We are the first company mapping over 1 million square meters that are all the Metro metro stations. So in some way, in metro station, we have met all for security reasons,
Steve Statler 07:40
For security reasons they want a map. And can you explain a little bit more about that?
Vincenzo Dentamaro 07:45
These are privacy policies that are commissioned to map everything. And they have policies running inside the Metro section. And they want to track that if there is something wrong happens, they can ask for a helper and synchronize that. And that's the basis. And what they asked us was to track the case is something like the police, but police.
Steve Statler 08:12
Okay, so you're tracking the employees and the police within the subway system? It's not tracking passengers.
Vincenzo Dentamaro 08:20
Yeah, no, I will not have the app that enables the, the tracking.
Steve Statler 08:30
Okay, and so who is the customer in that that case? It's the it's the metro system that are paying for it?
Vincenzo Dentamaro 08:37
Yeah. The public government that controls the metro stations, Commission, the main contractor, and main contractor is our technology to do all the stuff, recall.
Steve Statler 08:54
And so yeah, let's talk about what the boundaries are of what you do, because there's potentially a lot of software involved here. Can you kind of draw the box around the the piece of technology that you provide? Is it just a set of API's? Or do you have a GUI or?
Vincenzo Dentamaro 09:12
An SDK that can be integrated in apps? The SDK has the capability to localize the user indoors, and also offline, it means that everything runs on the smartphone. So there is no polling continuous polling on the server. The competition does not happen on the server. It happens on the smartphone, okay. It's because when you go some way you can also anytime we don't have internet connection, very strong subway, so we must ensure that everything works also when your internet connection is so bad.
Steve Statler 09:51
You're using static iBeacon UU IDs are you that this is basically you're not doing any encryption?
Vincenzo Dentamaro 10:01
Correct? Yeah, static IP customer ID that are not connectable. So they are there. And we are messing on the signal coming from that because we clean those kinds of signal and localize the user. Also, there are other things. Next to me is also accent. So official technology uses, we have a technology also called intelli walk that recognize and train some user step. So, when user is walking, if you are, if you walk with our technology, after 15 minutes, the technology has noted your motion model. And it's very passionate, I'll get your smartphone to another one, we'll need 15 minutes about 15 minutes to learn these motion model. And then we'll get you also you can disable the Bluetooth, scanning and walk. And it will take you all the steps and get you right on your direction on your position. Because it's very precise that it's called that recording. And it's very precise bend recording, because it is learning your motion model.
Steve Statler 11:15
So you're looking at the compass and the accelerometer. And you're using that to kind of figure out someone's motion, and then you use the beacons to kind of do cost corrections and figure out where somebody is?
Vincenzo Dentamaro 11:30
But with a major match together with Matt constraints, particle filters, and so on. And you've got a very accurate location. And also, if you are when you work, it prevents you from localized on walls and unaccessible areas.
Steve Statler 11:49
So what's the process of getting all of this working? Is there some kind of site survey that I need to do with an app to?
Vincenzo Dentamaro 11:58
The main thing is that when we started our company, we were using fingerprinting? Yes. First fingerprint, and then lab technology, simulation location and mapping, what did happen is, our customer were not able to fingerprint and even because also with them, you have to walk with normal acceleration, if you stop, you have to restart walking, you don't have to turn very quick. Otherwise, the accelerometers will give bad results. And that's very difficult to explain customers. Yeah, they are system integrators. They are not data scientists, they are not. You know, when they do that, also there are people who will do the mapping is even not a software engineer, a worker, so they don't have those kinds of skills. Right. the right skill they have is to mount a piece of hardware on the wall. Yes. That's all they know. Okay. In that way, we, our research was on doing filtering. So all unsupervised learning on the signals coming from becomes so we put the complexity upstream rather than downstream. And what happens is, is that they only have to attach the beacons on walls after the our suggested position. And, and the localization works.
Steve Statler 13:42
So you basically, presumably, like let's let's talk about the subway installation specifically. So you had how many beacons are there in that Subway?
Vincenzo Dentamaro 13:54
How many beacons? Well, remember, they will remember however, we we are covered for for instance, in also another honor another customer that is close, not robots. 10 12,000 square meters.
Steve Statler 14:14
112,000 square meters, 100 beacons and what what environment was that? Who was the customer?
Vincenzo Dentamaro 14:20
It was very noisy? The company's share price was revised, you know, they make cars see and there are a lot of robots working. So I run everywhere. Yeah, that makes no things very difficult for our filtering engine. Yeah, and it was working really, really really well.
Steve Statler 14:42
So we got two threads going here but they're both interesting. So let's finish off the Fiat one. So what what a Fiat using these beaconss for?
Vincenzo Dentamaro 14:49
Oh, they use the our technology to track the maintainer because what happens is sometimes robot I stopped walking. So they have only remember for five maintenance for all the rules, five, five maintainers for all plan, and the maintenance reach the robots that are stopping working, took over six minutes to do that. Six minutes in average, with our technology save about three minutes, because the technology calls the length trainer that is more close to the robot that stopped working. And they say 500 euros per minute. So 1500 euros every time the robot stopped working.
Steve Statler 15:40
That's a fantastic ROI metric well done. That's that's those very rare. So I'm impressed. So basically, you've got these guys and gals who are fixing the robots when they break down, and you're basically getting the right person who's closest to the robot that's broken down. And they're basically getting there faster. And so the factory stays running and they Fiat save a bunch of money. Correct. Fantastic use case. So let's go back to the so go back to let's go down underground, that the wrong subway. What was the process for deploying those beacons? So you were working with a systems integrator? What are you How did you? Did you kind of get a bunch of maps in the subway? And then kind of figure out? How did you figure out where the beacons we're gonna go? And how did you get the beacons deployed?
Vincenzo Dentamaro 16:34
And that's another interesting thing. It was not the very easy to, you know, to do all the work to our, to the system integrators because they have to learn about RF, radio frequency propagation indoors, that's quite a bit of, you know, a lot of work a lot of things to learn about physics, we decided to start developing an algorithm that synthetize all the knowledge about the RF propagation, and met with bad constraints to suggest where to put to the beacons. That's not very, very accurate. So maybe a human being will, and that has all the skills will do. Sometimes, that's not doesn't happen frequently, but can suggested in better position. But normally, the software can suggest good positions, to people that don't have those kind of skills.
Steve Statler 17:35
So part of what you have is software, presumably, this is kind of some kind of web package. And you give some inputs in terms of what the location is like, and it will suggest where to put the beacons?
Vincenzo Dentamaro 17:50
The user has uploaded the map, and the system automatically recognizes the walls. And you can also suggest where are other walls. And with those information, and knowing how many then I mean, one week on every, every out, if you want to go to one weekend, every 45 square meters, so you had to put fortified as parameters. And we understand to put one bigger every 45 square meters. With this information in the software provides the number of beacons. But with the number of beacons, it starts putting the beacons in the right position. And so finally, this is an AI algorithm. It's an optimization that as result will give you the map with the position of the beacons that the easily the you know the maintain in that case, the system integrator prints can give to walker.
Steve Statler 18:57
Then how do you make sure the beacons are in the right place?
Vincenzo Dentamaro 19:00
That's, that's the that requires actually a human or human being that has to see. Oh, that's okay. That's okay.
Steve Statler 19:09
So there's a supervisor that goes back and then that's it. There's no fingerprinting and there's no kind of walkthrough. You just work with that and then proceed.
Vincenzo Dentamaro 19:19
Is to give very easy up at the first real plug and play technology for an opposition.
Steve Statler 19:26
Very cool. So you provide the API's. And so far, you've got a deployment in Fiats you've got the deployment in the subway, it sounds like you've got a lot of blindness in museums, any museums that we would have heard of, or you want to?
Vincenzo Dentamaro 19:42
Oh, yeah. The states. You said you're where?
Steve Statler 19:47
I'm in San Diego.
Vincenzo Dentamaro 19:53
We have the office change gallery of San Francisco that has our system.
Steve Statler 19:57
The Art Exchange in San Francisco. Well, that's 500 miles North, but I actually do plan to go up there fairly soon. So I will, I will check it out. So they have an app that guides people around. Cool. And any other customer examples that you'd like?
Vincenzo Dentamaro 20:18
Yeah, we did. In 2014. In Italy, that is a big exhibition of wine in Italy. And we met at this exhibition was very huge. I don't remember how many because we use the one that was a 1000 square kilometers. We used 120 beacons 120 beacons about I don't remember the number correctly. But that was the number. And a lot of users were using the app. They released the app only for for for iOS.
Steve Statler 21:03
But what? So there's a there's an exhibition app. Can you explain the functionality that the beacons enable within the exhibition at Apollo the same way you are?
Vincenzo Dentamaro 21:16
Yeah, the beacons are only for us necessary for our technology to know when to use the position of user case?
Steve Statler 21:23
Yes. Okay. And so that's something that is that purely for the user to figure out where, how to find something. Or are there is there information that the exhibitors get?
Vincenzo Dentamaro 21:40
There are a lot of so the user could search and reach a specific exhibitor and, or showing the route the path through the reach the exhibitor and store the information, the phone number, shares, disposition to, to friends, to social network, also meet with other friends, imagine a LinkedIn, you get your connection, and you can ask your Lincoln in one of your LinkedIn connection to reach you in the position you are, you can share your position and they can reach you.
Steve Statler 22:24
It's always challenging meeting up with people at these shows. So what sort of what's, what is life? Like living in Southern Italy, I kind of envisage a lot of fantastic food and wine and great climate is that how do you get any work done? It must be tough.
Vincenzo Dentamaro 22:40
Oh, not right now day to the weather is not very, very good, let's say. But often, it's, it's good. And we also have, you know, good food. But food is not only the most, if not the most important thing in the life. And for us, as you know, software engineers. It's not the Italy right now the right place to live for that, but no.
Steve Statler 23:14
But so there must be some advantages and disadvantages of that. And, you know, I grew up in England and a long time ago, and when the when the gap The internet didn't exist. And so the gap between what was going on in Silicon Valley, and what was happening in London was huge. And now it's, it's much smaller. But you know, I think you're you also must have some benefits being in Italy, it's like I'm sure Italians would rather deal with someone from their own country to deploy technology, then go to someone in the states who may be is, you know, hard to get hold on, do you get any benefits from just the locality between you and your customer.
Vincenzo Dentamaro 23:57
There are two companies providing indoor positioning right now in Italy. And let's say, for what we know, the product that is really solely our product that is really to go to the market, as we do. And for this reason, we do have also customers in Italy. But the company was born with an idea to become global. Okay. And for this reason, we, we plan to open you know, other fetal quarters in California. We were, we were in Santa Cruz and San Francisco. And we will open in one to two cities, we will decide and another branch in Asia, where the markets were huge, too. We also have a lot of requests for that site.
Steve Statler 24:57
And so where are you at the moment in terms of none For people and funding that sort of thing.
Vincenzo Dentamaro 25:03
Yeah, the company is completely self funded. from us. We also had a grant from the, from the donation Italy, the grant is not so much 192,000 euros. And we are right now in fundraising. Fundraising, I mean, to searching for what in Silicon Valley, you call a seed in Italy is around a because number of different we are searching for for that because of the customers, we have the company's revenue generating. And the things because we have a plan, very complete plan that has been designed by our angel investor, Marco, that is also the president of the Italian startup, the delegation of startups in Italy. He's the president, and is also an angel investor, they as design form the roadmap for the ground. And we are searching also for an international investor that believe in our kind of technology. And so how do you earn money? What's the pricing approach for yourself? Yeah. And the pricing is based on license. So we license the technology on the squared meters. From our customers perspective, it's good, because when you go to other competitors, they price their license based on the users. So you will not know how many visitors your building will have. Right? So it's, you cannot, you will not know, in advance how much you will pay.
Steve Statler 26:55
Okay, so right, it says that you have a certain amount you offer certainty in terms of your pricing model. And I'm assuming, because you mentioned earlier that you don't need to be connected for this technology to work. And so maybe your pricing model is, as a result, if you don't need to be connected, then it's hard to charge by user. But do not lose something. If there's no connection, then I mean, part of the reason why people want to do indoor location and navigation is to spy on their, on their customers to to understand their customer better. Is that a disadvantage of your approach?
Vincenzo Dentamaro 27:38
Well, the approach is not completely a fly, the approach the approach, if you work offline, if you want to localize yourself offline using our technology, you can do that. But what didn't happen is that if the the developers that integrate our SDK wants to have more information about the user, he can actually do that. With there are few parameters in the SDK that allows you to get the position and transmitted the position, also, the positions that were not the start cache. So I mean, when you use it offline, your position our cash is a lot. And then this kind of log is sent through the network to internet when is available, and internet connection. So you get the same you know, everything about your your user, also when user was offline, let's say got it. But it's up to the developers to know, you know, the company that integrates our SDK to do this scandalous thing, and the visitor that has the app installed. must know that you know what he he's so the privacy is.
Steve Statler 29:00
Yeah, absolutely. Okay, very good. So I think you were looking for some metrics. And what did you find the metrics from that show or not?
Vincenzo Dentamaro 29:09
Yeah. We have 2531 downloads in two days. Okay. The app was in the first page, App Store. And we had a conversion rate. So I mean, who downloaded the app and using the app for over 30 minutes or so I have a bill that 32%.
Steve Statler 29:37
32% of the people that downloaded it.
Vincenzo Dentamaro 29:40
Used effectively using the amount of time.
Steve Statler 29:45
That's great. All right.
Vincenzo Dentamaro 29:46
Yeah. Because usually when you do see it, then you will install or close yeah, but in that case, it means that the app was very useful to find place.
Steve Statler 30:01
Well, that's cool. So next to me, I definitely very interesting technology sounds like you've got some great customers. interesting time in your company's trajectory, raising money and looking to come to the States very good. Anything else we should talk about?
Vincenzo Dentamaro 30:21
About funding, I can say that we decided to don't have early investors very soon because, you know, otherwise the valuation of the companies was very, very low. And the first evaluation is very important. And we decided to scrap our product with customers, big customers to get the first evaluation. Right. Well, I wish you luck on that. And it sounds like you've got some good technology and some good customers and you're working in an interesting area. So hopefully, we'll put this out and if we can be of any help and getting the word out with with this video, then I'll be very pleased. Thank you. Good to meet you both. Ciao. See you. Thank you.