Learnings and Outcomes of the Use of a Digital Travel Credential (DTC) Type 1 for Boarding and Border Crossing Processes, a Pilot Project in The Netherlands

Thank you for having me here in Berlin. The last time I was in Berlin, Roger Waters performed the wall and you know then how long I have been going, I was a student then. Today I am working in the Ministry of Justice and Security as a civil servant and architect in the program of border innovation and something that was really in the heart of our program. We did in the first quarter of this year receiving travelers from Canada by when they had provided their DTC upfront. Now the digital travel conventional, Annette mentioned already a little bit of what it is, so I'm not going deeper into that.

We worked with that and we worked with that because European Commission sees a potential in that for innovating border control. And that's the main cause of why we're doing this. We see a potential and a potential needs legislation in order to become real, but to do the legislation, taking into account everything that that yet you run into when you're doing this. They wanted to try it out, try it out in practice and that what, that's what we did. Not only should this lead to insights on legislation and directives, it should also help travelers. It should bring travel facilitation.

So we wanted to look into that. We did that and we have some insights about that.

Of course, it should still be safe, safe for the passenger, his his data, but it should also be safe for our national safety for that no one crosses the border who shouldn't be crossing and aviation security as well. Part of the pilot was using the DTC for boarding in Canada and next to it, because we did TRA travel facilitation, we also experimented with the questionnaires on how long, but for what purpose and with what means you are going to stay in the sheen zone. This was the first time, at least in the Netherlands, that we did it in a digi digital way.

So we, the traveler wasn't quite questioned on that when arriving in Amsterdam. Now who are we? I said a lot about that. I'm from the Ministry of Justice and Security. This ministry has the final responsibility for border control in the Netherlands and we had partnering ministries as well. The Ministry of the Interior who is responsible for issuing travel documents was involved. They were very interested in digital travel credentials and how that would work in combination with passports. And the third ministry is the ministry for Ministry of Defense.

The Royal Netherlands Marsh you say, is the main agency enforcing border control in the Netherlands. And they do so, especially at Shri Hall Airport, we wanted to invite travelers. Now how do you do that? The best way is to involve an airline. KLM was involved. They approached Travelers Forest and asked them, will you please participate in our pilot? And the next partner was RI Hall, the airport Amsterdam.

He, they gave us the facilities to set up things. EMIA was the technology provider. A picture paints a thousand words. So what is a video? I think it's the best way to show to you what we did. This video we made to inform the public about what we were doing At the request of the European Commission. The Netherlands is conducting a pilot with A DTC, A digital travel credential. The DTC one pilot investigates whether registering at home for boarding and border passage makes processes at the airport more effective and efficient.

The results are input for a new EU regulation for a digital travel document that the European Commission is working on. The Dutch pilot is a collaboration between the government, KLM, Royal Dutch Airlines, skippo and idia. If you're traveling with KLM from Canada to the Netherlands and if you're over 18 and have a Dutch, Belgian, or Canadian passport, you can participate in the pilot prior to your trip, download the DTC app register by reading the chip on your passport, you take a selfie, which is compared to the photo in your passport.

You then share your relevant data with KLM for boarding the aircraft and with the Royal Netherlands Marish shall say, for border crossing at Skip Hall, Because The marish shall say receives your passport details in advance, they can carry out most of the necessary border checks before you arrive at the border. At Skip Hall, you'll then cross the border faster using the special DTC tap and go border gate. Your DTC will be retrieved via a facial scan. You then hold your clothes pass board against the border gate. A physical passport is therefore still necessary for this pilot.

If there's a match between the DTC and the passport and there are no specifics from the previously carried out checks, you can cross the border. The DTC contributes to a smooth process for the airline, the border control authority and the airport and to a pleasant journey for the passenger. So this is basically the overview. Travelers traveling in the first quarter for this year with KLM from Canada to the Netherlands, invited to join the pilot. When they would join, they could download an app that was in the app stores of Google and and and Apple.

And with that they could derive the DTC in the way just shown by reading the passport, by making a selfie and getting the DTC stored on the phone. Next they were enabled and asked to enroll. People traveling from Montreal Airport could enroll for boarding because that's where the mon the gate was.

And all the other destinations in Canada were able to enroll for border crossing in Amsterdam, both when arriving in the Netherlands and when passing onto other areas in the shingen zone in between that's not in here in the sign is that there was a pre-assessment of the traveler's information by the Royal Netherlands. ma you say they did this by hand because the pilot was not connected to border systems, but there were the same checks done as done on the border when you arrived there.

And in addition to that, they did this evaluation of the questionnaire people sent in about their means purpose and and length of stay. Now there were interesting learnings from that. Let me show you the gate crossing. These are the actual gates that were in there. The left one, it was in Montreal, the other one was one of the gates in Amsterdam. And next to it you'll see the process. What happens on this gate, the moment you approach the gate will automatically take a photograph because you have enrolled your passport photo from the DTC is already in the gallery of that gate.

We'll try to match you and the selected DTC at that moment, the gate will ask the traveler to present his passport. It need be opened just it can be offered closed. And with the data in the DTC, it'll try to open the document. They'll open the chip of the document and perform an active authentication to establish that. It is indeed the DTC that was used, oh sorry, the passport that was used originally to create the DTC from. And then you get something that goes like this. This video is six seconds and 0.93 and that's about the time we can achieve with this process.

No border checks have to be done at this moment. They have been done upfront with reliable DTC information. Same process applies to going through the for boarding. It's the same way of establishing identity in numbers a little bit, and I'll get some details through these numbers as well. We invited 16,000 people to join the pilot. It was close to 10% of the people started the process by downloading the app and well for an email campaign, that is, that is quite a good response.

And then you see the first area where we had a little bit of trouble and that was the trouble in creating the DTC people needed to derive the DTC from their passport. And there were two hurdles in there. The first was actually reading the chip. People had an average of two and a half times they needed that to read the chip before it went successful and some failed, some didn't come through that. And then the second thing is that we had a liveness detection in there, which I believe is important. But also the other half, about about 300 people failed in that as well.

I must say in there are also 300 people who started but were not eligible. We had interest from Iran and Uzbekistan and a lot of other countries, but they weren't eligible. Another interesting thing, which which really well wasn't, was better than our expectations. The Canadians were asked to fill out the questionnaire, which was quite a large job. It took about nine minutes for them to do that. But of the 600, 364 people, sorry, four 11 people that started it around 90% finished it. So there were, well even they were willing to do that in order to do the, the border crossing.

It takes some time to deriving the DTC and enrolling was around four minutes and the questionnaires around nine minutes. But generally people were willing to do that and that was very good. Good to see now on the way the numbers in Montreal are lower because only people flying from Montreal could be in there. And then on the second last part, there is something that we really missed a lot of people on the airport because all other exits were still of course available to travelers. A lot of people just walked out to the other exits and we didn't see them on the border.

That was a pity, but, but, but it showed, it worked. What did we learn coming into that?

Well, the most important thing is we really put ourself in a hard place by saying we want to test the DTC with real passengers, real life situations that mean we had to cross all the boxes be for quality control of privacy, of human rights and algorithms of of information security. That all got got implemented and it actually works and it is actually a lot faster and has enormous potential for increasing processes at the border. As you see, the fastest time we've recorded was 6.5 seconds. Of course that was a traveler or someone who was really instructed, who knew exactly what to do.

The average of of of 12 to 14 seconds is I think a reliable data we got, we got from this. And this is about twice the speed that we have today with our gates at at RI airport. So this is really an enormous improvement. As I told earlier, usability is key and we are sure we can improve on that.

The, for if anyone of you have tried it, it is a little bit cumbersome to get it, to get it done even after a lot of design. And we need to, to work on that to make it really easy. And of course for the future, we did it separate from current processes on the border that goes not for boarding because you can't do boarding separate from, from other systems. So that was connected, but it was separate from border control systems would make it easy from a pilot management project management perspective. But however, for for scaling up, it needs to be integrated.

But furthermore, it needs to be integrated in the infrastructure on the border control filters, it's Ripple airport. And so it should be a common process where you automatically get in the right flow when you have this. Something about the app. The app in the pilot we did everything because we needed to do everything. We did the derivation, we did the storage, we did the enrollment for KLM, we did the enrollment for kmar all in one app and and web app technologically seen.

When you put everything in one app in in really tech, technical and app, that would be the ultimate possibility to make it user-friendly. However, from a a future perspective it would be more logical to separate it and to have derivation and storage of the DTC in one place. And there's a lot of discussion here in this, this, this arena about how doing that right and where it should be. So I think that's, that's better addressed there. And then have enrollment in different apps for these different service providers.

That's, that's makes it a good separation of concerns and very logical management. However, one other thing, at least in our pilots, we chose for storing the DTC on the user's phone, not on a central storage. That in in our opinion, gives him maximum control over what he does with it. He can delete it and be sure that it's gone. I know about also the security risks attached to that. I'm sure that we had a very short lifespan. So that needs first the discussion in the future. But that's was our, was our choice.

We also minimized the data we used and therefore we overdone it a bit because we also left out communication options with the traveler. And that was a little bit difficult. For example, when we had to inform them that the gate wasn't in operation for at a certain date and that was something we need to change. Next slide we'll be on border control and I'll go then more, we'll talk more deeply about that. But somewhere in the process we need to do a thorough passport chip inspection.

And then I think one of the, the, the main learnings that we can take from, from our pilot is that it will be very difficult to create an app that will read all passports on all phones. We had about seven types of phones and three passports of three countries, which is only a very limited number. And we had a a well very big effort to get that all running. And that will be, so it will be something that, that that won't be available for everyone when you use it. This technology only it it'll be, can be done for a large crowd, not for, for everyone.

Then on border control, I said earlier the, the willingness to answer the digital questionnaires, that was really great to see. We asked people to upload, for example, a hotel voucher to to to give some evidence about their means of stay for the permit and they, they did that and that allows for border control to go two ways. It's digital so you could do it more quickly and, and even digitally.

But on the other end you have, when you get it earlier, the opportunity to do it more thorough, you could call the contact they've given up that to verify that indeed the person is going to meet the person he says he's going to meet and both options are available. It's quite too early to say what we'd like to do with it. But that was very good to see. On the other hand, there was an, an interesting thing that happened.

These questionnaires were checked by border guards and these border guards had became very formal in checking that information they said, well I I'm, am I convinced that indeed you have the means? Well I can't read the hotel voucher.

So no, and they rejected the and that's because you've taken away the traveler and the, the, the border guard seeing the person in front of him. So we need to take that into account as the future. It is a hundred percent digital process we did, no one looked in the booklet and for future situations we need to look into that because this needs to be enabled in EU directives in legislation because it's either explicitly or implicitly assumed that you're doing that. So for what travelers, in what cases is this indeed a good way?

We're doing a hundred percent of digital information and that means that we need to be very sure about that digital information on the border. We do very thorough chip chip checks. We the ArcHa O standards, whether they are are, are really followed. Do all the signatures, checkout is all the information consistent and that should be done in there. We did a lot in our app because we fully control the app, but if we can't fully control the app or cannot be sure that it's actually done, we feel that it still must be done at a border crossing point at least once.

And that's something in, in the design we came as an important issue for, for us because, and that's only very simple. The trust in the DTC is only as high as the trust in the passport. It is derived from, well on the last thing we need a permanent directives in place to, to enable us to process that traveler data. We need to know from one and up till one we can check that information. Those were the learnings. There is a, a great perspective for this and well thank you for hearing me out.