Silvija Seres: Hello and welcome to Lørn, my name is Silvija Seres, our topic today is biometrics and my guest is Tugberk Duman who works for Futuice, a very exciting innovation company from Finland that we've spoken with on several occasions, and you're their head of innovations.

Tugberk Duman: Yes

Silvija: Tugberk we are going to talk about biometrics, basically the mystifying, the term and talking about the social consequences of it. Before we do that, can you tell us very briefly who you are and what brings you to biometrics?

Tugberk: Sure, very happy to be here. My name is Tugberk, I studied in Finland six years ago and during my studies I've been in several projects. One of which was a project for trophic wardens, their problem was checking the cars for the tickets and it was a painfully long process.

Silvija: These are the people who go around the parking space and check that your ticket is still valid in your window.

Tugberk: Exactly, so the problem was, it was a painful process which involved several different devices and several different API's, so what we've done, we simplify the process with a Google glass, actually it was the first commercial app for Google glass, and Finland till you just look at the camera, sorry the license plate, camera snap the photo, you recognize the license plate and check if payment has been done or not. And you are able to proceed with the penalty or proceed with another car?

Silvija: Too efficient a solution if you ask me.

Tugberk: Yeah exactly, that was fair to say that people really didn't like it but traffic wardens really enjoy the efficiency of it. So it comes with thinking then what if we take the same approach to removing the friction between human and machine interaction in general. So making it so efficient by just looking at things, by just being present we can actually authenticate you, we can give you access to whatever it is that you want access to. And that’s what brings us to biometrics, and that's how I started investigating biometrics which led me to establishing the biometric business unit in Futurice in the year 2016.

Silvija: So what is biometrics?

Tugberk: Well how I define it is, it is a unique measurable characteristic of you, either its physical or behavioural, so physical is facial recognition for example your face, iris, fingerprint, voice that relates to, that answers the question what and who.

Silvija: So we have to stop just for a little moment. So iris, there will be a camera, it will take a picture of my eye and there is something unique about the blue area in my eye, with its patterns that can, is it a one to one recognition, globally or?

Tugberk: So how biometrics usually works is that it's basically pattern recognition, it's a subcategory of machine learning, so what it does is, it looks at certain landmarks of your face or fingerprint as well. Certain landmarks like face recognition algorithms will look at two hundred different landmarks of your face, your eyes, your lips, your looks.

Silvija: Compare the size of my nose to the size of my mouth to my cheekbones.

Tugberk: Exactly, and same goes for iris, if you look into certain landmarks in your iris, all these tiny, the cameras are able to see it, then it turns into a mathematical model. At that point basically what we do is next time you end up in front of a camera or a fingerprint reading machine, you basically compare the model that we had before to the model that you are presenting at the moment.

Silvija: So basically you would, my face will become a sequence of numbers and the next time you look at me with your camera, if the sequence of numbers resembles the first one, then it's a close match.

Tugberk: Exactly, that's how it will basically compare two models that you presented at different times.

Silvija: So there is a lot of computations and the reason why a phone can recognize you quickly enough to open your phone just based on your face even though you might, I don't know, does it work if, when you shave off your beard in your case or I put on a few pons.

Tugberk: Yeah it does work, these models are quite well trained and quite ready for you know superficial changes and physical appearance of course, if you are talking about face recognition, if you get an aesthetic surgery, that's a different case but overall these algorithms provide you a probability of you being given a certain time. So for example if I take a photo of you today and put it in an algorithm, it will register you and it will say that this is Sylvia Seres right, and tomorrow you come, you gonna stand in front of the camera and it's gonna say this is a 95% chance that it's Sylvia Seres. Three years from now you will come, you will present yourself in front of the camera, the probability will go down because in three years you have gone through certain changes. So that's how it works, over time the probability will go down based on the changes you have experienced over the years and if we're gonna get there may be in detail but the way to keep up every now and then is to refresh the database with a new one.

Silvija: We have to take pictures of people, often verifiable pictures. So you are saying that this is a paradigm shift in a way and what you mean is that now these things are suddenly so useful and so available that they are going to change the way we live our lives?

Tugberk: Yeah. I think it's part of a bigger trend, what I see is that we are being thrown at oceans of information every day, why our phones, email, social media and our attention span is  getting lower. I've read somewhere this might be an intended phenomenon and I'm not sure but it said that our attention span is Lower than that of a goldfish nowadays.

Silvija: Meaning 30 seconds.

Tugberk: No 7 seconds.

Silvija: 7 seconds.

Tugberk:  7 seconds on average, that's how much attention we can pay at a time without being distracted with all the information that is thrown at us. So what biometrics does, these technologies does is, they remove the non-essential steps from the process, so what I want to if I give an example from unlocking for your phone, I want to reach my email and writing down a pin code is a non-essential step that I have to do. Now with your face you don't need to, you just lift your phone and it recognizes you.

Silvija: Open magic.

Tugberk: Yes, exactly. So that is the paradigm shift I'm talking about, we are gonna go to a point in the future where we don't see any gatekeeping anymore. So you just enter places, you just access places and the environment reacts to you rather than you introduce yourself.

Silvija: So you have one of your own examples for this, that's the face recognition payment, and then you also mentioned to me that you think the amazon goal is an interesting example, can you please tell us about both goals.

Tugberk: Let's start with the face recognition payment, payment process is already quite streamline nowadays, you have the NFC cards, you have the Google pay and the mobile pay, but what we are aiming for with project of that was personally involved in, can we remove the friction aka the need to carry any device or to gather, so you will just be a hands free going to shop, getting your purchases done and leaving.

Silvija: Can I just stop you for a moment, because when you talk about these near field communication, NFC payment systems and even things like apple pay et cetera, what you have to do is pretend you have an item that's  like your credit card and you have to wave it somewhere close to some sort of payment reading solution. You are trying to remove that step, we shouldn't have to hold anything that pretends it's a credit, we should just be ourselves, go into a shop, take what we want, leave and the system should be able to charge us correctly for that.

Tugberk: Exactly, complete naked payments if I may call it. So what we are working on is basically removing the mere friction of you actually running an email and you end up in front of a cash and you need to go through the apps to find the mobile pay app that you have for example. Or you don't have your wallet with you for one reason or another. So that's what we have done, we have a tablet that is acting like a point of sales device next to cash register, of course you need to register beforehand, take a selfie of yours, attach your card number to your face and when you present yourself in front of the tablet, you clicked pay it face, it recognized you, the cashier put the amount that you just needed to pay and the amount was transferred to the cafeterias' bank account.

Silvija: Could you combine these amazon goal like technology? Tell us about Amazon's goal.

Tugberk: Yeah, let's start with the amazon goal, amazon goal is one of the most prominent examples of the paradigm shift that I told you about, so what they do is in retail, they completely remove the need to queue having to present your card, go to a certain checkout system. Again if you remember what I said about removing the non-essential steps, when we shop, these are the things that we have to do, we don't enjoy waiting in the queue, we don't enjoy actually presenting our card, but it's just something that we had to do. So what they did is remove all these non-essential step, again the same logic you just reach to your face by a selfie, by a touch of a credit card number, you go into the physical shop, and then you

Silvija: which is an amazon goal shop

Tugberk: Amazon goal shop yes, so it's not any shop but amazon goals shops and then you basically go in there, there are cameras everywhere, there are smart shelves that understand which item you are grabbing and basically they match your face with the items that you grabbed from the shelves, and they automatically reduce it from the card, deducted from the card that you put in the system. And then you will without checking out, or waiting in the queue, just walk out.

Silvija: You just go.

Tugberk: Yes.

Silvija: So I guess there is some sort of a smart camera staff around these shelves that know the items and can match them. And this is really a disruption for retail that people I think underestimate because basically what happens is that once you learn that in amazon goal shops, you never have to queue, people will start shopping on amazon goal shops even though maybe they can charge a few percent more than other shops, and then of course the margins will get even worse for others and so, but this is not easy to do for a traditional retailer that has spent generations of people optimizing for the flow in the shop in the completely different setup.

Tugberk: For sure there are a lot of legacy systems that are in place that is not easy to change or to upgrade to this kind of technology to be broadened.

Silvija: So the one step of friction is that you have to register when you go in to an amazon goal shop and that is because the want to remove the ambiguity of marching your face to the exact credit card.

Tugberk: Basically it's a, you don't need to do an exact before they visit the shop, you can do it any time before and you do it once. So what you do is go back to the technical details a bit more, you basically tell the facial recognition system hey, this face belongs to me and I am this person, and this person has this credit card number. So when you go there the system as I told you, the system have a model of you that you present it beforehand and then you're at the shop you have the second one and then it compares to your thumb and sees that this is Sylvia and she's grabbing a banana so I'm gonna reduce one US dollar from her bank account.

Silvija: So you're trying to do a similar thing with e-shops that are not amazon goal shops for payment.

Tugberk: Yes not exactly but similar, different use case where we have tried was a cafeteria, so it was more of a grab and go type of setup, so it wasn't like an amazon goal physical shop where you have things on the shelf but rather there is a counter you put your coffee and you grab your banana or sandwich and then you pay and checkout. So what we have done is, instead of using your card or near field communication payment method, we put a tablet in there that act like a point of sales. Device you use that one and go. So overall what we have removed there is friction to case something physical but also when you have it, the fact that you need to take yours to detect that you need to get your card, you need to put it back, overall we are able to save three to four seconds from each payment compared to near field communication by just removing the need to present any...

Silvija: ...and the other thing is that I never walk out of my home these days without my phone but I have regularly forgotten my wallet, and so, where do people go to learn more about these kinds of new solutions?

Tugberk: What I do, you usually is I try to look at the big paradigm shift, I don't personally follow any technology could develop up just because there is a technology at the core but I try to understand how they fit into big paradigm shift that we see nowadays, and for example what we are following at Futurice nowadays is zero user interface friend, which biometric actually enables, so you remove the interfaces. So I will recommend people to get into understanding what zero user interfaces are, how we can change products and services to be compatible with it and there are quite a good amount of resources on the Intranet.

Silvija: So such a zero interface.

Tugberk: Zero user interface applications and services.

Silvija: Very cool, if we just spend one minute talking about the big elephant in the room which is really personalization versus privacy, because when you know my face, you can track me anywhere, and if we put this into my home we get even more useful because you know it can do all kinds of things, it's me but again then you know even more data about me. What should we think about this things or what's your personal take on this controversy?

Tugberk: I mean if you look at the bigger picture, what we see is a certain threshold where people start prioritizing convenience and personalization of a privacy, like the fact that we give iPhone or to any android phone, our fingerprints, our iris', our faces just to unlock it, it tells something about our prioritization in our daily life itself. So how I see this is that there is a way which we can make possible by having this convenience element, we are actually compromising from the privacy aspect of things, and that is what in my daily work I am aiming to. What I see in the future is that when we have this multiple face recognition system, multiple biometric system, voice recognition et cetera, we need to give the control of this personal sensitive data to its owner. So what I see in the future that you own your personal biometric data, are you gonna go to concert tomorrow, you wanna access with your face, you share it with the concert organizer for tomorrow for a certain period of time and once the need for the use is ended, the biometric data returns to your phone and the concert organizer don't sit on your biometric data forever. So that is the only way I see biometrics will come, it’s gonna change the way we interact with these intelligence services, but it needs to be on our terms, on people's own terms. So people own the data and people give it to the service provider on a need to know basis.

Silvija: Need to know basis is an important thing here, and I think also the responsibility that comes with you know providing this super comfortable services, because I think we haven't been quite honest with the users, and so we were saying while they signed that I can use their data and I need the data in order to provide them with the service but I think we've known for a long time that convenience will triumph personalization without people, sorry privacy, without people really understanding the cost to privacy. And so helping them to understand that we are very responsible and this is where the borderline is and this is how we will use it on a need to know basis will be a wonderful trend.

Tugberk: I think that's the only way feasible.

Silvija: Would you like to leave our listeners a quote?

Tugberk: Of course, this is not coming from a very commercial person, a talk leader globally but it comes from one of our customers that I personally work with, we were talking about how biometrics is going to change the world but we were also observing that in Finland and in Europe there's a certain hesitations around the technology and he came out and said surrounding world is never ready but waiting is not the best strategy. And I totally agree, once we have the chance, we can actually start doing something and set the right practices for deploying biometric technologies for the rest of the world.

Silvija: I would actually go one step further, by the way OP bank is really a cool bank, and I will go even one step further and say that waiting without knowing what you are waiting for is never a strategy. That is just hesitation. What will people remember from conversation if there is one thing they could take away?

Tugberk: Can I say two?

Silvija: Yes.

Tugberk: First thing is, biometrics is usually associated with surveillance and security, it's gonna come off from a different direction, it's gonna come from the service business, so it's gonna change how we interact with different services and different products for sure in the near future. And following that, the second thing is, if companies who have any customer interfacing product or service are not investing and investigating in biometrics today, they are already too late. So I will recommend getting on with it as soon as possible.

Silvija: Tugberk Duman from Futurice thank you so much for coming here and helping us learn more about biometrics.

Tugberk: Thank you very much, it was a pleasure.

Silvija: Thanks for listening.