Latest / Autonomy Insiders / Auve Tech’s Driverless Shuttle Playbook for Airports and Beyond | Taavi Rõivas

Transcript

Auve Tech’s Driverless Shuttle Playbook for Airports and Beyond | Taavi Rõivas · Autonomy Insiders

• Daniel Abreu Marques: Welcome to Autonomy Insiders, the show where global industry leaders in autonomous driving unpack their real-world insights. Taavi Rõivas: pretty much every car with this yellow light blinking that you see at the airport can be replaced by our autonomous shuttle Well if you have an airport, if you have a corporate campus and you as the owner believe that this service is safe enough, You can give us the permission instantly. You don't need to wait for law being changed, which means that this part of autonomous driving can scale very, very quickly. Daniel Abreu Marques: Welcome back to Autonomy Insiders. most autonomy coverage is about robotaxis in San Francisco, in Phoenix, or in Wuhan. But today we go somewhere else. So Estonia has 1.3 million people and has no big car industry, yet one Estonian company builds driverless shuttles that run in small Japanese towns, in downtown Florida, and since August, also on the airside of Tallinn Airport with nobody on board. And this company is Auve Tech My guest is its chairman, Taavi Roivas. before tech, Taavi was Prime Minister of Estonia, the country that puts its whole government online, and today we cover who Auve Tech is, why they pick low speed shuttles, when most money chased robo taxis, what the business really looks like and how the deployments are going. And also what comes next. So Taavi welcome to the show. Taavi Rõivas: Thank you for having me. Daniel Abreu Marques: So Taavi you run a digital country, then you choose a company that builds vehicles and hardware is always the hard, slow part, also the capital heavy part of tech. So why did you pick the hardest part for your next step? Taavi Rõivas: Building the vehicles is a small part of what we do. most importantly we do design the vehicles and we d make them drive autonomously, which is the most interesting and innovative part. and for me, Auve Tech has always been a company that has the potential to change a big part of our lives, the way we commute. Yeah, and I do believe that we are very close to actually achieving that. so I'm very happy that I decided to join already six years ago. Daniel Abreu Marques: And what did running a government teach you that helps you today at Auve Tech? And maybe also what habit from politics did you have to unlearn when you joined a startup? Taavi Rõivas: Well, a lot of habits I had to unlearn because you know there is a saying that how do you know you are no longer a minister? Well you sit in the back of your car, buckle up and nothing happens because you no longer have a driver, right? So in startup you need to keep things lean, you need to be able to work on all sorts of topics and and that's actually something that makes it also interesting. Additionally, of course, what is the really exciting part is how you can really do something that can transform big things in the world, including the commuting. Now, what is the biggest lesson I can use from my political life? Probably that the Moore's law doesn't apply to regulators or politicians. it doesn't become easier to regulate How do we allow vehicles to drive hundred kilometers an hour on public roads a site or next to human drivers? So still, even there where you have permission to drive on on highways autonomously, most regulators, if not all, still want to keep the human in the loop. And and that's one thing that we at Auve Tech can understand and believe that will not be changed very soon and our business model relies on having a human in the loop still but only a small part of his time so he can concentrate on several vehicles and that's what makes the service actually economically profitable. Daniel Abreu Marques: maybe now let's dive right in. So many listeners probably know the name Auve Tech but not the story. So maybe let's start here at the beginning. Taavi Rõivas: Uh-huh. Daniel Abreu Marques: So could you explain what Auve Tech is doing? So what do you sell? To whom do you sell? Taavi Rõivas: Well we were founded already in two thousand eighteen, actually before that we launched the first prototype but already road legally driving on public roads since twenty twenty, so quite some time. and there is tens of vehicles driving every day on public road, most of them in Japan, but you mentioned also Florida, also Abu Dhabi, some European countries. But now we have realized that in addition of driving on public roads and showing that the technology actually works, the best way of actually growing and making an impact is to take the safety operator away from the vehicle, which is easier and makes more sense on private roads and use cases like airports, like big corporate campuses are ideal for that. And since a few weeks back we have demonstrated also publicly that we are the first European company that has the permission to take safety operator out and take passengers on at the same time. And that is the kind of precondition that makes the service competitive so that we can really like offer something that has been done previously for transits or human driven shuttles. Daniel Abreu Marques: we will also get later a bit more into your deployments, but before we go into that, so what I also found interesting is that Auve Tech came out of a university project and is now owned by Silberauto so a car distribution group. So how does a car dealer end up owning an autonomous vehicle maker? So what does that owner give you that venture money would not, for example? Taavi Rõivas: Mm-hmm. Well, first of all, it's Silver Auto Group is not just a car dealer and the car dealership actually is no longer in the in the group. So there has been thirty years of experience of designing and building custom made Mercedes, for example, like ambulances, police vehicles. So that gives us the automotive DNA and and know how how to actually effectively and Wisely manufacture vehicles so that the small scale can also make sense, not only in thousands of units. Secondly, it's a lot about the founder. He's himself very hands-on in first of all designing the shuttle but also giving the young team that has created the autonomy, has created the mechanical parts. giving the young team a chance to create something that is world changing. So that is the basic synergy of having a corporate or family office ownership instead of VC. Having said that, of course, we are also exploring the kind of VC market to see if at the verge of expediting our growth it would actually make sense to to invite some VCs on board as well. so we would not outrule VC in investments for sure. Daniel Abreu Marques: Makes sense and the robotaxi race got billions, so a lot of money flowing in there. And when you look at shuttles there were players like Navya and EasyMile. And Navya went into receivership in twenty twenty three and easy mile moved also into industrial towing. So why is slow speed shuttles a good business in your opinion when some of your best known peers struggled? Taavi Rõivas: Mm-hmm. Well, first of all, I think the main reason why they struggled is that they were a little bit too early to the market. the technology was not quite mature yet, all the lidars computing power, all that. ten years ago when they first had their operations, it was not there. and they never reached this One safety operator per several vehicles, which will make the service economically sensible and affordable for the fleet managers and end clients. having said that, I think both Navya and EasyMile have done a great job in creating the market as such, introducing autonomous last mile shuttles to the world. Now, why this segment is particularly interesting is the the reason that on closed areas, there is already hundreds of thousands, if not millions, of last mile shuttles that rely on human drivers. So the demand is is huge. Even though it's not as big maybe as robotaxis it's still huge. But what is our advantage in comparison with robotaxis? Well if you have an airport, if you have a corporate campus and you as the owner believe that this service is safe enough, You can give us the permission instantly. You don't need to rely in at least in most countries, you don't need to rely on regulators catching up with the technology. You don't need to wait for law being changed, which means that this part of the autonomous driving can scale very, very quickly. also if we think of autonomy as an industry, it actually makes a lot of sense to start with smaller form factors and slower speeds, not trying to regulate hundred kilometers an hour driving lorries to the highways immediately. So demonstrating the technology at those environments where it by definition is safer, that makes actually a lot of a lot more sense. Daniel Abreu Marques: Absolutely. And you said you build both the vehicle and the self-driving software and most of the industry says okay, pick one. But can you explain why you want the full stack at your size and is this also still the plan going forward? Taavi Rõivas: Well interestingly enough, there is thousands of companies that do something in autonomous driving. But if you want to have robot taxi, you have a few options. if you want to have last mile shuttle driving autonomously it that actually works, you also have only a few options, maybe five if you count the Chinese companies. If you don't count the Chinese companies then maybe like three companies that actually can legally drive on roads. so for us saying pick one doesn't really make sense. you can pick one but then you don't have anything to offer to the humankind. You just have some theory that you can scale in the future. But we are very practical people. We have the shuttles driving on the streets of of Sakai-machi or Abu Dhabi or West palm Beach and we see that the you know this form factor which is a small eight seater with our technology stack, with our autonomy stack, actually provides the service. so we have everything that is ready to be offered to the client. Daniel Abreu Marques: Makes sense and maybe now let's open the hood and look into your vehicle more closely. So walk us through the MiCa shuttle. You mentioned eight seats, it also has I think seven lidars ten cameras, redundant steering and braking. Can you explain a bit why you chose this design and also maybe which design choice cost you the most? Taavi Rõivas: well if you remember the first modern electric vehicles, they were built on internal combustion engine vehicles and just modified, which was not optimal choice, right? The the best modern electric vehicles are the ones that are built to be electric from ground up. The same is with autonomy. Of course you can take Jaguar, of course you can take whatever robot taxi companies use those OEM vehicles but they are not optimal for autonomous driving because they have driver's seat, they have steering wheel which is wasted space for an autonomous vehicle. we took like tabula rasa or clean sheet of paper and designed the vehicle from ground up to be autonomous. Actually it was designed inside out because our founder who did the first drawings and the concept, he realized that the best way of fitting eight people to very small form factor, let's compare it with Toyota Yaris, for example, is to have people sitting opposite to each other and then have folding seats on the side as well. we can fit eight people in the form factor of Toyota Yaris of course it doesn't need to be aerodynamic, so we can have this kind of boxy shape offering more shape inside, so it's more like a bus than a passenger car. and lastly, things that really make sense in autonomy is redundancy, double braking sensors, double steering sensors, double communication. So think of it as like a two-engine airplane. even if something happens with one engine it can still land safely. the same is with us. even you know it's not very likely of course that the steering controller fails, but in our vehicle the likelihood is exponentially lower still that because you have several controllers that that can work independently. Daniel Abreu Marques: And you also launched MiCa 2.0 in twenty twenty five. Can you also explain what changed technically to the first version and also what the vehicle Taavi Rõivas: Mm-hmm. Daniel Abreu Marques: can now do what it could not do before? Taavi Rõivas: Well MiCa 2.0 is two separate things actually. It's the change in the form factor that visually looks very similar, but the actual design changes are in how to manufacture and assemble the vehicle. So making it much faster to assemble, much more effective so that we can easily take part of our assembly to United States, to Middle East, wherever the market is. but second part of MiCa 2.0 is of course the software stack that is a lot more capable, that can do a lot more things including maneuvers that were not available in the first stack. And making it like very simple, the level of autonomy needs to be close to hundred percent. If not hundred percent maybe In some cases it's ninety nine point something because there might be some edge cases. and based on this mathematics, we can calculate how many vehicles can one safety operator remotely monitor. And and if the vehicles get stuck for some let's say I don't know, garbage truck blocking the way or or whatever kind of edge cases, then the person that is remotely monitoring can say that wait or can say that you know now you should find a way to to pass the garbage truck. so yeah it all comes down to autonomy, the percentage of how big part of the route can the vehicle drive by itself. Daniel Abreu Marques: Makes sense and if we now think about a new route so how fast can you open a new route today and maybe also what is the real bottleneck with that? Taavi Rõivas: Mm-hmm. Typically we do it in twenty-four hours to a couple of days. this means that we have 3D mapped route, we have possibility to do it with the vehicle itself, but usually how we do it is we send a special device that is put on top of a normal passenger car, the route is mapped, and when the actual autonomous shuttle MiCa arrives. the map is already being treated or or like yeah prepared by our team. having said that once we have the initial setup in place we usually would prefer doing the operations with a safety operator on board for a certain amount of kilometers. It might be a couple of months even. and after we have received enough data that the vehicle has been operating for thousands of kilometers on this particular route without any issues, only then we advise the client that now is the time to take the safety operator out, and then we agree with the client how many vehicles one safety operator can safely monitor. So again, coming down to the mathematics and the mathematics is not based on theory. But in real data that that we collect from this actual use case. Daniel Abreu Marques: staying on the remote operator theme. So in the US you work with Guident for teleoperation, but in Tallinn you run remote supervision yourself. So is in your opinion remote operations a core skill you must own or is it also later a commodity that you just buy? Taavi Rõivas: in Tallinn we don't develop the technology ourselves. So we use external technology, we just use our own remote monitoring personnel for now because the Tallinn airport case is a landmark case for us. This is the first where we actually took the safety operator out and we want to learn as much as possible from it. So so the person actually monitoring is our engineer. but having said that, in other use cases we prefer doing it with third parties or with fleet operators. We don't want to have our people in each and every location. But luckily, yes, there are companies that design that part of technology so we can rely on them and the service or technology is good enough that we are happy with it, so we don't need to develop it ourselves. By the way, this was not the case with the vehicles. in the beginning of our company we looked a lot like which platform would make sense to be used for autonomous shuttle, and there simply wasn't any optimal vehicles. You can have yes, OEM vehicles and these are a little bit easier to get legislated for the vehicle itself, but they are by no means optimal for actually offering the service. So We didn't want to compromise in this and that's why we came up with MiCa, which is the ground up autonomous shuttle. Daniel Abreu Marques: the good tech alone is not a business so I think we should also talk about the money and how the money works for your company. So how do customers pay you? Is it then buy the vehicle? Do they lease it? Do they pay per hour or per kilometer more as a service? So which which model do you push? Taavi Rõivas: And so far we haven't done any kilometer based or hour based services. We either sell the vehicles and offer monthly recurring services which will be software updates, which will be support for whatever service and repair. we also have full service rental operations. But I'm pretty sure that this all will change once we will show the fleet operators of this world or large fleet owners of the world that if you take depending on market, but if you take at least three vehicles, we can assure you that during the first year already you will see that autonomous fleet will be cheaper to run than the human driven alternative. And then basically it's up to them if they want to purchase the fleet, if they want to lease the fleet, the same as they do with their old school minivans, they can do with our vehicles and just have the saving from the workforce. So they need a lot less people to be employed and that will create a kind of economy for them. Daniel Abreu Marques: maybe we can also come back to the remote operator theme because you also mentioned that it's a part of the economics that is really important. And I think today in Tallinn you have a remote operator ratio of one to two. So at what ratio does the MiCa beat then also a minibus with driver on costs, so to speak, per passenger or kilometer or something like that? Taavi Rõivas: Well, first of all, there is no secret that our vehicle will be at least today, more expensive than conventional minivan because we have a lot of technology on board, lidars cameras, computers and so forth. but that doesn't really mean a lot because the saving will come from the most expensive part of the service, which is the person driving the vehicle. And we have calculated that in most markets if one safety operator can monitor three vehicles, it is already preferable in comparison to human-driven vehicles. in Switzerland, for example, where the salary is very high, it can be only two vehicles monitored by one person. In Dubai, where there is a lot of affordable workforce it might be five, but it is definitely in the range that is technologically doable. So technology wise again it depends on the use case, it depends on the ODD, but three or five is not the limit. this is just the breaking point where we can be competing in market terms with the human driven shuttles and if we can raise the ratio further which we can then of course our service becomes increasingly attractive and even if we are you know at the same price with a conventional service we already have an advantage because in many countries where we operate there is difficulty to find the people to drive those last mile shuttles. Let's face it, it's not the most exciting job to drive back and forth, for example, between a hotel and airport terminal or or inside the airport where It's always the same scenery, you drive like five kilometers there, five kilometers back. So finding people to work there not the easiest. so but yeah, having said that, we are aiming for better value for money as well, not just matching the price of conventional human driven shuttles. Daniel Abreu Marques: And a another part of the economics is also utilization when you talk about autonomous vehicles. and I could assume that probably the utilization is lower if you, for example, are in an airport setting than with a robotaxi service. So could you explain what limits the utilization maybe in your case? Is this demand, route length or something like that? Taavi Rõivas: Mm-hmm. Well, I think the only limitation actually is the fact that we have electric vehicles and at some point they need a bit of charging time. Other than that, we can operate them twenty-four seven. And the question of course is do the clients need it twenty four seven? At some points they might need a Tallinn Airport, our client is Magnetic MRO All, which is actually a company that operates during office hours more or less. so for them it's enough if we do seven days a week and I believe eight or ten hours per day. but of course if the client requires to have longer hours the vehicle is not the bottleneck for sure. only the charging time, but that would be for every electric vehicle the same. then of course having electric vehicles. We still consider being a huge advantage in comparison with IC Daniel Abreu Marques: Absolutely. And this also means that the longer the customer wants to operate, the more economic your shuttle gets for them because at some point they need probably two driver for one shuttle or something like that and you can also just do it with one. Taavi Rõivas: exactly. the ratio that I explained that the breaking point is if we have two shifts a week. But if we have four shifts a week, which would be the twenty four seven service, then pretty much in every market if one person can remotely monitor two vehicles, it will be already better priced than having a driver in every vehicle. So really like we are very, very, very close to the breaking point and it's just a question how to get quickly to to all the relevant markets. We know that in the US alone there is approximately two hundred thousand human driven last mile shuttles only on the private side of the road or like privately owned roads. So that of course is an immediate market opportunity for us. Daniel Abreu Marques: And also you also mentioned that your bus will probably be a bit higher in cost than the traditional shuttles. So could Taavi Rõivas: Mm. Daniel Abreu Marques: you also share what share of MiCa's cost is the additional sensors and compute that you need to Taavi Rõivas: Mm-hmm. Daniel Abreu Marques: to perform the autonomous driving task? And maybe also how fast that is falling, because you also hear a lot about lidar prices falling and something like that, at what Taavi Rõivas: Mm-hmm. Daniel Abreu Marques: price point maybe also opens up the mass market. Taavi Rõivas: Yes, that's a very good question. but I I don't think it's a deal breaker anyways, but but also for robotaxis, the technology stack costs probably more even than the car below it, right? In our case it's might maybe like one third, maybe forty percent. I don't know for sure, but but that's something like that. and of course it's it's decreasing and also we see that if we produce no longer tens of vehicles but hundreds of vehicles, we can also reduce the vehicle cost further. But again, even if the vehicle costs more than let's say Ford Transit or Volkswagen Caravel, we will still make the most of the same from the most expensive part of the service. And in most countries, in most cases the most expensive part of the service is the human behind the wheel. so that's where the actual savings will come from. And of course it's our job to show the clients that even if the vehicle is slightly more expensive, you should look at the overall cost and there we will perform better. Daniel Abreu Marques: Absolutely. if you look at traditional buses, there's a huge driver shortage because you don't have enough drivers. This is also the same for for airport settings. Taavi Rõivas: It is it is definitely the same and it's maybe even worse because at the airports as I said before the driving routine is very very basic or it can get very boring for any human being. Also at the airports the drivers need to go through the security tests or clearances, so that takes additional time on onboarding them. because there are special procedures and you need to make sure that that you are vetted or like checked. so yes finding those drivers to work there is a huge challenge in most of the markets at most of the airports we have been talking to. So we are actually solving a real life problem. Not just making the solution more affordable But actually solving the bottleneck for the airport owners, but not only. companies that are present at the airports, having locations in different hangars need to move their people working there. pretty much every car with this yellow light blinking that you see at the airport can be replaced by our autonomous shuttle Daniel Abreu Marques: And now maybe let's move from the spreadsheet to the streets. So where are these vehicles running right now and also how is it going? So your first driverless service in Europe runs at an airport, as we now talked a lot about, and also we talked about why you chose that airport first. But what did the regulator need to see before they said yes to you that you can operate without driver Taavi Rõivas: the regulatory of course was familiar to us before because we had done a lot of POCs on public roads, but there they always have required safety operator on board, which as we explained before, has made the service economically pointless. it's nice to show that the technology works, but cost wise it will never outperform the human driven shuttle service. and secondly, the regulators or decision makers want to see actual data. So show them that you have actually driven on this particular road safely for a certain amount of kilometers and that convinces them. It's a lot about mathematics, a lot about statistics or probability, and if they see that the probability of something happening is very, very low, then they have the certainty of giving the permission. Also when we have done our operations and are doing our operations on public roads, in every country we run a lot of tests. I think the most difficult tests we have done are in Japan. our vehicles were tested for half a year, I think there was like between 60 or 70 tests that were done like driving in darkness, driving in rain, driving on slippery slope, what kind of obstacles do we detect, how small does the detection go, all of this. and after six months of testing we got level four permission in Japan as well but again to actually scale it to actually take the safety operators out in most of the use cases It's so much easier and so much more logical to do it on private roads. It will happen on public roads as well, but I believe the regulators will take a bit more time to allow it. And and every safe kilometer that we are operating is bringing it closer, but it still takes a bit of time. Daniel Abreu Marques: what I also found interesting is that at the magnetic MRO the shuttles does not only move workers but also parts and equipment between the hangars And I think you probably expected that it carried mostly people. So what surprised you about how the customer actually uses your shuttles? Taavi Rõivas: Actually if you come to think of it, it makes a lot of sense that you save the time of the person working at your factory by just using an autonomous shuttle of sending one box of things, it might be tools, it might be you know, parts of airplanes, whatever, you just send it with an autonomous shuttle and you don't need to ride along. If you have a driver or normal passenger vehicle. you need to drive along, right? Or you need to steer the vehicle. so saving people's time has become a very interesting by product of our service and yeah if you come to think of it makes sense that people use it this way. Also we are seeing that moving goods, moving parcels autonomously makes a lot of sense as well. so we are definitely not outruling to exploring this market as well. But our main focus today is to be best in the world in this last mile shuttle service that takes people from point A to point B. And once we have achieved that we will definitely look into other verticals as well. Daniel Abreu Marques: you report a ninety-nine point nine percent autonomy rate. So so what is the zero point one percent? what what happens in that zero point one percent and also who handles that? Taavi Rõivas: Typically this is the remote monitoring person giving permission that yes it's safe to go. In some cases we have this double checking by human being requested by the client. For example when it is crossing taxiway where airplanes can be as well. Of course, our shuttle detects the airplanes, detects we have done tests even with hovering helicopters, detecting them three meters from the ground and so forth. But sometimes if the client says that you know here it is mission critical that there is like double checking by a human being as well, then we just create a stop point there and remote operating person needs to give a go ahead as well. And that will be the kind of zero point one percent. now having said that achieving ninety-nine point nine percent in the airport is a lot easier and makes l a lot more sense than achieving it in let's say city center of Mumbai where there is million people and and walking on the street. so I think For all the companies developing autonomous technology, it makes sense to look at those use cases where autonomy actually functions best, not to necessarily try to do it in the busiest city centres where there is a lot of risks and the ODD is exponentially more difficult. Daniel Abreu Marques: what what's also interesting for me is that you have really different kind of use cases right now on the road. As we mentioned, you have Japan, you have Florida and you have Tallinn, for example. And I could assume that you have more older residents, for example, in Japanese towns, you have this visitors Taavi Rõivas: Yeah. Daniel Abreu Marques: in Florida, you have technicians at an airport. So can you share a bit about how the acceptance to the technology differ between those categories? Taavi Rõivas: it's actually amazing how it works in Japan. There are number of small cities where there wouldn't be possible to organize conventional bus service or public transport service because the cities are too small, the streets are too narrow, there's maybe not enough passengers for fifty seater bus. But the small eight-seaters function perfectly and I have taken a number of rides all around Japan with old ladies, with children, travel from point A to point B in an autonomous shuttle. And I can tell you, like in the beginning of course it was an exciting thing. Today it's their routine. so they don't think of this shuttle Daniel Abreu Marques: Mm mm. Taavi Rõivas: necessarily as as something that is autonomous. they see the shuttle as enabler to go and see the friend that lives in the other part of the city. Otherwise it wouldn't be possible because the walking distance would be too long and they don't drive their own cars anymore. So you are very correct to say that in Japan there is a lot of elderly people to whom we offer a new possibility to commute. And it's really great to see that some Japanese cities, for example Sakai Machi that was the first that introduced autonomous transport in Japan, they see that this is the best solution for their people to be able to move to every part of the city. Daniel Abreu Marques: And staying in Japan, I think it Japan is also one of the hardest car markets in the world and you said you took forty eight percent of Japan's autonomous shuttle deliveries in twenty twenty three. So how did it come that a smaller Estonian company wins over there? Taavi Rõivas: Yeah that's a thing we are really proud of because Japan is the biggest car maker in the world and then selling cars from Japan to Estonia should be much easier than selling Estonian cars to Japan. But maybe it's because the OEMs still are looking at passenger cars mostly. maybe it's because you know Mercedes has not built an ambulance by themselves they rely on companies like Silver Auto Group to do it for them in in markets like the Baltic States or Nordics. so probably it's the same in our niche. maybe they see it it's still too too small for them to develop and that gives us the opportunity and then Once the market is visibly big enough, well then let's see if they will come to the market and compete or maybe they want to cooperate with us. so that remains to be seen. But today I wouldn't say there is a real competitor from Japan arising or existing. there are companies that are actually partly together with us or in partnership But that are developing their own software, but yeah, I don't see any vehicle or like complete service arising from Japan yet. Daniel Abreu Marques: And you now have dealt with a lot of regulators. So Japan, as we mentioned, Estonia, you also have the US and also the Gulf. Can you also share a bit about so which was the hardest to deal with and also maybe do you have something that Europe should copy from other regulators? Taavi Rõivas: Well for an Estonian company getting your autonomous shuttles road legal and ready to be sold in Japan, US, Middle East and all around Europe, that's a thing that we still can't believe has been possible because it's an insane undertaking, of course, or a huge achievement. having said that every country is difficult in their own way, and everybody has slightly different regulations. Also within Europe we have more than twenty different regulations it's not all the same at EU level. but of course we have also been able to use the previous acceptance from different countries to get the further in the other ones as well. So of course every track record brings new trust. having said that probably Japan is the most difficult again I should ask the team to make sure but yeah half a year of going through tens of tests that was really something and yeah if you can make it in Japan in the very difficult car oriented country then hopefully you can make it everywhere else as well. I would say that also US of course is is quite a challenge for a European company, but so far we have managed it really well so Daniel Abreu Marques: Sounds good. And I think in twenty twenty three you signed with Abu Dhabi, Saudi Arabia and also Qatar. So did those turn also into paid fleets or are there more like demos over there? Taavi Rõivas: in Abu Dhabi we have actual operations which is like a commercial project together with Masdar City, which is the kind of innovation hub for Abu Dhabi. in the rest of Middle East there has been a lot of interest, a lot of POCs, a lot of testing, a lot of demoing, but the actual commercial traction has been slower to become. but having said that, as we speak, we have our CEO in the Middle East and talking to the big names there who potentially would want to scale the fleet also in this region. So we definitely have high expectations also to Middle East, but also we continue to be bullish in Japan. And we also believe that because US has so many human driven last mile shuttles, the same actually is in the Middle East, they have a lot of last mile shuttles that are human driven, these markets should be actually the most logical for us to pursue. we don't need to like offer them anything new. We just need to tell that okay, your existing human driven shuttle costs you this much annually if you have a fleet of five or six or nine. then our offering will be more effective throughout the fleet and you should opt for autonomous shuttles instead. Daniel Abreu Marques: Makes sense and maybe now let's end with looking to the next few years. So where will Auve Tech be in twenty-four months? So maybe also in how many vehicles? Maybe also which segments carries the business? So is it an airport or have you been gone to public transport or something like that? Taavi Rõivas: I believe in twenty-four months it is likely that there will be first operations without safety operators on board also on public roads, but the real scaling has happened during that twenty-four months in private areas such as airport air sites, commercial campuses, These kind of areas where the owner of the land, owner of the facility can give the permission based on the data that we can provide, and we don't need to wait for the law to catch up with the technology. Daniel Abreu Marques: And also you have Chinese players coming to Europe and they have robobuses, maybe also even lower prices, and also big balance sheets. for example, WeRide is also at Zurich Airport So how does a European company compete and is also made in Europe here a real sales argument, or is it more like a slogan Taavi Rõivas: made in Europe definitely is a sales argument. Designed in Europe is even bigger sales argument and being compliant to GDPR and basic hygiene in cyber security and data deployment is the biggest argument. There is several countries, including US, Japan increasingly also Middle East of course also in EU that is skeptical about Chinese technology coming to our markets mainly because they have suspicions how the technology is using the data that is collected. We have read a lot of cases by some of the Chinese companies that probably influence the others as well. not in a good way. So having said that, I think that Chinese technology is worth looking into. It's definitely I wouldn't say that they are not doing a good job in autonomy. They are. But yes, the the main factor why why you should look at European companies or American companies is the way we we treat the data and how transparent we are about it. Daniel Abreu Marques: maybe a last one to close. So you used to write the rules so and if you could change one EU or national rule tomorrow to scale driverless the shuttles, which one would it be? Taavi Rõivas: The biggest bottleneck today for autonomous driving on public roads is the need to have a safety operator on board. Once we have the rule that several vehicles can be remotely monitored, then I believe the sky is the limit. There is plenty of last mile shuttles driving with human drivers on public roads and private roads alike. On private roads we are ready to scale already, and I do believe once the regulators see that this is actually a safe thing and needed for the society, they will give us the permission to drive without the safety operator on public roads as well. Daniel Abreu Marques: So we're at the end, Taavi Thank you so much for taking the time today. It's was really insightful to learn more about Auve Tech. And yeah, thanks for sharing all your insights and knowledge and happy to had you on the show today. Taavi Rõivas: Thank you very much.