Latest / Autonomy Insiders / Inside Loxo: Driverless Delivery Vans in Europe

Transcript

Inside Loxo: Driverless Delivery Vans in Europe · Autonomy Insiders

• speaker-0: Welcome back to Autonomy Insiders. I'm your host Daniel and today we have a special episode. Since we were recording live from the Automated Mobility Summit in Zurich and I took the opportunity to grab the CEO of one of Switzerland's local AV champions. So today I'm talking to Amin Amini He is the co-founder and CEO of Loxo. And it's also special for me because I took one of my first rides in an autonomous vehicle yesterday in a Loxo vehicle and it was a really great experience. So Amin, so happy to have you on the show today. Daniel Abreu Marques: Today with Amini. Welcome to Autonomy Insiders, the show where global industry leaders in autonomous driving unpack their real-world insights. speaker-1: Thanks for having me. speaker-0: Yeah, sure. So let's start with the company itself. So I think in Switzerland, almost everyone knows you probably also in Europe, but maybe outside of Switzerland, Europe, not so much. So please give us a rundown on Loxo So what is the company? And yeah, what are you doing at Loxo? speaker-1: Happy to do so. We are a company ⁓ a level 4 autonomous driving solution ⁓ to the logistics and commercial vehicles. So we don't do people transportation so far. We are having a laser focus on the logistics ⁓ cases for the moment. And what we build ⁓ Loxo is the complete ⁓ software stack. And we work with the different partners, integrators to... make our stack available for transportation companies in the sense that we are providing a complete solution for those who are having goods to transport. Currently we are having a big focus on the urban and sub-urban areas, where the scenarios are a bit more complex and the speeds are a bit lower. Yeah, so we are doing that since now three years on the public road and yeah, you're right. Loxo is a known name in this environment. speaker-0: and also trucking gets a lot of attention, but maybe the, as you mentioned, urban goods delivery, not so much. So walk us through the strategic logic. So why did you choose that kind of vertical? ⁓ And also why Europe? Why now? speaker-1: from a strategic decision, focusing the ⁓ development and that specific use case is mainly driven by the market opportunity. Because if we look, we have great companies doing ⁓ heavy trucks, robotaxis But even at the worldwide level, we have only few companies really focusing on that type of application. but then when you down the view on Europe itself, then the number of choice are very, very limited. And then we saw that as an opportunity from the business perspective. And then we said, okay, we have to build a solution that is answering the requirement of that market. know, if you look to this whole European market, for example, from logistic perspective, trucking is a very big market. Because at the end of the day, are a software company, so we are always looking to the number of cars available at that segment of the market. If you look to these heavy trucks, we are talking about more or less 6 million heavy trucks in whole Europe, which is great. And if you look to the light commercial vehicles, those type of vehicles that we use, typically we call them delivery vans, we have around 30 millions. Therefore, the market opportunity was quite appealing to go in direction. yeah, the other reason is that we were always searching for impact. You we just don't want to position ourselves in the automotive chain being just another supplier and another supplier and so on. We were searching also for impact and we thought that with the direct solution to the transportation companies, we can create more direct and tangible impact in a short time. ⁓ speaker-0: Yeah, makes a lot of sense and maybe give some more context the company. So how big is Loxo today? Also, maybe ⁓ how many vehicles do you have? speaker-1: Yeah, basically we are a team of more or less 30 ⁓ people inside of the company. As I mentioned, we are on the public road since 2023. So we have been in Swiss by our customer, Migros which is a big name in the retail industry here in Europe. And we have doing ⁓ pilots ⁓ already back 23. From there, we scaled ⁓ to cover an entire city, a logistic application. This is something that we do with partner, Planzer which is also a local logistic company, it's very well known here. And we have built a lot of things around this project, which is running in the city of Bern, the capital city Switzerland. And last year we expanded also to Germany. So we have now our, since a few months, our first experience also. with the European markets and Germany is a big market for us. So this is where we are currently. What we do is commercial pilots with our customers, which means we are not for the moment focusing on increasing significantly the number of vehicles. This is a ⁓ stage to a bit later. But what we are doing is taking all the market requirements from regulatory perspective, from the business case of the customer. and making sure that in these commercial pilots we can hit these targets. And this is where we are quite efficient and for us in terms of revenue generation we are just the starting the recurring revenue. This is where we are. speaker-0: Sounds good. Yeah, we will get to the different of you later in a bit more detail. But what was also super interesting to know is where do you see Loxo operating in some years? So will it be then only Switzerland and Or do you also envision ⁓ some day operate in the US? Or what is your ambition there? speaker-1: Well, for the moment we are having a strong focus on Europe obviously. But in the meantime we have ⁓ quite some interesting requests from the ⁓ manufacturers actually, ⁓ always in the commercial sector. So have OEMs in pipeline that we are working with already. And those OEMs international. ⁓ So it's not that we are going to... expanded naturally by ourselves, by our own fleet all over the place. through these OEM partnerships, we are going to deploy our technology in multiple markets. And obviously, U.S. is one of them. When Let's see later. speaker-0: And I imagine you cannot say which OEMs it will be. No. But super interesting to hear. And then do you also talk about ⁓ the ⁓ about and factory installed solutions so that ⁓ solution gets ⁓ and the factory line or is then more like a retrofit kind of partnership that you've started? speaker-1: Basically for the transportation because we have always these two types of customers. mean those commercial vehicles that they have their own application, their own use cases, there we are the tech provider. But when it comes to the transportation, we are ⁓ having a ⁓ with of the largest European car manufacturers that is going to produce these vehicles according to our specification on the production line. And then to answer your question, if it's like the retrofitting or is going to be the complete transformation of the car, we see it as a process. So we are going to start step by step this collaboration. And once the market is also getting and we are to transform even deeper at the vehicle side. But in all case, we don't ⁓ go to car manufacturing directions ourselves. We see ourselves as a tech provider. speaker-0: Makes sense. And maybe let's also go to the vehicle side since we're already talking about that. You run several vehicle platforms, ⁓ as I read in my research. So we have the Loxo Alpha that you build yourself. You have the R1 retrofit and also the VW ID BUZZ Can you walk us through ⁓ how this mix came up and what is the logic behind that? speaker-1: As I said at the beginning, we are mainly focusing on the logistic operations. Basically, for us, ⁓ the question of the two years ago was not the most important thing because we had to prove the technology is... working well enough to hit the KPI's of the customer and also the authorities. Because at the of the day, what we are about to do is to remove the safety driver from the steering wheel and this is something that they can disclose. therefore we went with the solution which the component of the market are more or less ready for such activities and you know Volkswagen and ID.Buzz from Volkswagen is a vehicle that is used already in multiple contexts. And at the same time, we have ⁓ Amag ⁓ and Noviv as a partner, which the importer Volkswagen here in Switzerland. So they kindly offer us to use that vehicle. This is how we have started. What we realized and what learned is that, yeah, for some use cases, we need significantly larger vehicles. So this is something that we are building now. speaker-0: sense and you also touched on the topic of safety drivers so one question I always like to ask is do you have a timeline when you want to remove safety drive already speaker-1: ⁓ Yes, we have. ⁓ Yes, to remove the driver from the steering wheel, it's a of days. ⁓ speaker-0: Okay, that sounds really good. really cool. forward to hear that. ⁓ speaker-1: Yes, we do. Yes, safety is really our highest priority. So we have been lucky enough to be to Swiss authorities to build a very safety-oriented plan ⁓ for removing the driver. So we have been receiving ⁓ very specific KPIs we have to achieve. then after this achievement, KPIs, we are to remove the safety driver from the steering wheel and put it on the passenger seat. ⁓ This is a matter of days. for us, ⁓ the vehicle, it's on the road since 2024, September. And in the last ⁓ months, the safety driver, which is inside of the car, didn't touch the steering wheel. So is something that it's important to mention that, we can see also different means of the KPIs that this ⁓ presence of the driver behind the wheel is not anymore necessary of the maturity of the technology. We put it in the passenger seat, we have to prove again the same that yes all the safety features are are very well present, that we don't have interventions and later on we can completely remove it. speaker-0: Can you also tell us a bit more about which KPIs you need to hit? So what kind of KPIs do have? speaker-1: This of course, for example, often ⁓ if any and the safety intervenes. Because obviously if an ⁓ ADS system still a safety driver to intervene, this is not level four, right? So this is what we had to prove. This is the main KPIs. But as we have also this remote supervision, the quality of network, also classical things about Okay, how smooth is our car in the whole traffic? Also from the social acceptance, are the people happy about this project or there are some complaints from the citizens? So we have to prove that multiple dimensions of autonomous driving are met in our project through removal. It's not only about the tech, of course the tech and safety are the most important part, but there is also the requirements which are going beyond. These are the main things that we have to monitor and report. speaker-0: makes sense and you also talked about different projects that you have already and the safety driver out is in one specific project or everything that is the drives in Switzerland. speaker-1: We start with one specific project which is the one in Bern ⁓ Planzer ⁓ this is the starting point. speaker-0: Okay. And can now walk us also ⁓ bit to difference between the projects? You mentioned you have a project with Migros, also like really well company and now also with Planzer What is the difference between those projects? speaker-1: Well, basically when we talk about ⁓ focusing on the transportation of goods, are talking about a countless number of use cases. So this whole market, the most important aspect is, okay, ⁓ what are most important use cases to start with? And here it's also an opportunity for the transportation companies to transform their ⁓ process. Actually, you look to this industry, The things have not changed that much in the last 80 years basically. You know, we had always the same concept of delivery. Maybe now the big transformation is that changing the thermal vehicles with electrical cars, but then the concept of delivery itself, there is not much a change. So now with this autonomous driving technology, we are exploring completely new ideas, new use cases. ⁓ that could improve the efficiency of those delivery processes. So the difference between your different projects is that we are experimenting ⁓ the different commercial pilots, the use cases. And now after doing that for three years, we have a clear idea what are the use cases that makes more sense and ⁓ we have to go, ⁓ with whom have to go. This is becoming more and more clear. For example, we have started with Migros, we have been delivering goods, the goods actually, from Migros to the campus of Schindler, which is a big lift company. And that was about the convenience. Then if look about the project we have Bern with Planzer this is about the transformation. Actually, we are doing a completely new use case. We have divided this whole chain of delivery with middle mile and the last mile. And yeah, we see a perfect fit of our solution for this middle mile, like a shuttling between the different points in a dynamic way. call it very often a meet and feed, the last mile delivery persons, which was not really available before this technology because it's also economically makes much more sense to do it with autonomous driving. And if look to Germany, we experiment purely B2C deliveries. We know all the challenges, but we wanted to do that and to learn also about the opportunities, but also the challenges. So this is a bit our strategy with different customers, with different partners. We are trying different cases to the sweet spots of the technology and the business. speaker-0: Makes sense. And touched on the Planzer use case and mentioned the middle mile. So how long is usually ⁓ middle mile segment? So how many kilometers do you drive then between the hubs? speaker-1: This is ⁓ really depending the size of the city. ⁓ Planzer is... They are very strong with real estate. They have very well depots across Switzerland. In Bern, we are talking about few kilometers, less than 10 kilometers. But if you look to the bigger cities at the Swiss scale, ⁓ Zurich, then maybe we have to add maybe double it. But then the same scenarios if you look to the US market again is much bigger. why we have to keep the flexibility also from the vehicle size because the vehicle size also change based on the geographical situations. But concept is valid everywhere more or less. speaker-0: and now we know what Loxo is doing in the different project. But what's also always interesting is what is the business model? How do you then envision to generate revenue? So are you selling the vehicles then? Do you lease them? Do you charge per delivery per kilometer? So yeah. What do you envision there? speaker-1: Well, all the learnings that we had in different projects, we came to the conclusion ⁓ the way is to not change drastically the business model of the transportation companies. If you look to the way that they function, it's quite they purchase or they source the vehicle from a partner and then they hire themselves the drivers. So then we said, yeah, why not doing the same? So let's... them source the vehicle from our preferred OEM partners, OEMs that are integrating all the sensor and doing all the modification that we need for the car so they can purchase the car. And then instead of hiring drivers, they can hire digital drivers by the means of our software. So the business model is the quite identical for the transportation companies. And from the Loxo perspective, obviously we are talking about recurring revenue. So we have per vehicle, annual ⁓ per vehicle. This is how we are charging our customers. speaker-0: Makes sense, also helps Loxo to keep light. Correct. So you don't have to have all the capex on your balance sheet. Absolutely. Super interesting. Yeah, so we also now know the business model and I now want to get more a bit on the technical side. So let's open the hoop a bit. So the Loxo driver is described as two components, LOXOFuser and the LOXOGuard working independently. Can you explain us? a bit as a why decide to split them as well and how it works. speaker-1: Basically, the industry of level 4, a lot of companies are going with the AI-first approach, which means basically solving any challenges or any functions by AI, which is a great strategy and we really like that. the reason is that it's very, very scalable. But since day one, we knew that we are going to have the European requirements for putting the vehicle on the road. So therefore we said the AI first approach alone is not enough. So we came to the conclusion that yeah, we cannot expect from the same type of software technology, take this end to end transformer based learnings and models. We cannot expect to bring all the necessary performance while being fully compliant with all the safety, cybersecurity, performance norms and standards that we have here in Europe. Therefore, since day one, we have started the system architecture by these two components. One component that it's really bringing the performance. This is the one that customer needs to be efficient in their process. And in parallel, we are coming with a second component, which is ⁓ a safety guard, let's call it, we call it actually a Loxo Guard, as a co-pilot that supervise all the decisions that are made by AI and how to make sure that we are always remaining in the boundaries of safety. And at the same time, this Loxo Guard must be certifiable. So, because we cannot just put black box, we have a lot of rules and regulation here in Europe. that they are asking us to do that. And this is part of our company culture. We are having a very strong safety culture. So we wanted to make sure that you are not only talking about the performance of the ADS system, but for us, the performance is first safety. And then it comes to the whole driving functions. So this is the idea behind to make sure that we have the full traceability to all the applicable directives, rules, regulations, standards. ensuring the certifiability while not making too much compromise the performance basically. This is the main reason behind why we have this double component system architecture. speaker-0: So do you then also envision that other companies that are using this end-to-end approach that you mentioned already have then significant problems to come to Europe because of that, because it's not that explainable? ⁓ speaker-1: Basically, first of all, mean, this a science segment that is developing very fast. Two years ago, explainability was a big issue, but now it's becoming also more and more feasible to explain how the things are working. If it's enough for a full certification or let's call it type approval, for the moment, I would say it's not enough yet. The science needs to be developed. But to answer your question, those companies that you're focusing purely and only on the end to end, if you look, they are also mainly looking for level L2, ⁓ plus, L3, which means ⁓ is a safety driver or a driver. So if you are targeting pure automotive value chain and you want to become a supplier to an OEM for level two, ⁓ sometimes we call it level two plus or level two plus plus. then yes, it is possible and we don't need to go further because in the safety case we can say there is always a driver to take over. But in our case, we are not targeting ⁓ level 2 or level 3. We are targeting really level 4 for logistics and commercial vehicles. Therefore, we have to start with that since they want. speaker-0: And you also talked about a different kind of approach to mapping. it's called virtual mapping and ⁓ is an to the pre-built HD maps. Can you also walk us through that? speaker-1: Yeah, so basically this end-to-end model doesn't consume any maps. So that's actually the big advantage of such end-to-end models. But actually we are having a reference to these virtual maps as we call it for our Loxo Guard. So the second component which takes the safety responsibility. again... we have always a commercial view on such technologies. For example, if you look to the price of the, call it high definition maps, so the mapping per kilometer, and then you look to your customer business case, the number of deliveries that they have to do to only break even the mapping, I'm not talking about the whole hardware, the process change and the software training and so on. Then we see some challenge actually, because we are, very very oriented in our decision from the customer perspective not only for our own benefit but what makes sense for our customers therefore we came to the conclusion that this is not a commercially viable approach so we had to find early days alternative ways and we found an alternative ways that we call it virtual mappings which means that We don't necessarily go on the streets with a fleet of the cars fully equipped with sensors and the post-process everything because it costs a lot. So we found a way ⁓ we can do it by sitting in the office and using multiple sources of information and doing that virtually. And then we go on the streets only for the validation, which is ⁓ significantly ⁓ effort that you have to spend on. And you can automate that. So this is something that we have built. Therefore, in our business case with our customer, the mapping is not anymore a bottleneck. Expansion of the whole TOD is not anymore a bottleneck because it's not always about, okay, now we have to send the fleet of the car for mapping. No, it's not the case. So we see the advantage on the business case of the customer, but also the scalability. This is probably even more important because you can deploy this map. significantly faster. speaker-0: Sounds really good on a cost side and maybe let's stay also on the customer side. So in trucking, ⁓ especially long haul trucking one thing ⁓ we always is driver So it's a huge problem, a huge bottleneck everywhere in the world. Is this also the main reason why your customers want to go to your solution because there is also a lack of drivers for the shorter routes and for the urban areas? speaker-1: Well, yeah, of course there is also driver shortage. If you look, there is a lot of turnover of those ⁓ urban drivers as well. are lot of people going to retirement. And I think the statistics tells us out of four newly hired drivers, ⁓ one leaves the next 12 months. And this whole driver shortage challenge on top of increase of the parcels, you know, because everyone orders more and more online. So these online orders or even the B2B volumes are increasing while the number of drivers is decreasing. But to answer your question, actually, I wouldn't say this is the only and the main reason. This is one of the main reasons, course. But the transportation companies, they see also this technology a way for improvement, for making the whole thing more efficient. And it's not always about the cost or always about the driver shortage. Of course, there are very, very important components in the whole discussion and the whole decision of a customer. And this is given. mean, this is a challenge that everyone has in urban delivery. Yes, we absolutely have the same type of challenges. But great of such a partners that we have is that they don't ⁓ have the focus on driver shortage or cost. They see, no, this is a way to make everything much more efficient by changing the whole process because now everything is possible. You can extend the delivery range hours. can ⁓ think about completely new use cases. You can get very closer to the customer. You can give much more flexibility to your customer. So this is the way that they see this technology more Of course, solving the driver short range while improving the costs. speaker-0: let's assume everything works well and then five years from now, where do you envision Loxo to be as a company? speaker-1: Well, we are really set to become the solution for the logistics sector here in Europe. So this is something that we are really focusing on. We are onboarding very key partners onboard from the customer side. We are on those ⁓ as well that we can be faster in deployments where the regulations are bit more prepared, like here in Switzerland, for example. In five years time, would like to be the solution ⁓ for these logistic companies for European ⁓ ⁓ So this is where plan goes actually. And at the same time, we are enabling already and we would like to increase this OEMs customer base. and there... let's see what the automotive industry in Europe will bring us in the coming years. speaker-0: So you want to, plan to have like a multi OEM setup more or less, so not like partner with one specific OEM, but have like broader OEM setup and the customer can choose basically whatever OEM he likes most and then buy the vehicle speaker-1: Absolutely, that's clear. ⁓ have also, I mean, the second part of what I said, we have also the OEMs as a customer. That we purchase the vehicle or our customer, don't, but they have their own business cases or their own customers. And this is a customer base that we would like to also enlarge. And we have ⁓ quite ⁓ demand the European OEMs. And the fact that we are a European company, centrally based here, It's quite convenient actually for them and if you look to the readiness of the ADS for European markets, we can say we are quite in a good position as well. especially if the use cases are on the commercial side, then it's even a better match. speaker-0: And you mentioned you're a European player based here and it's convenient for the OEMs, but we also see a lot of, ⁓ for example, Chinese players now entering Europe as well, of the time on the passenger side, but could also be that they want to go into logistics sometimes. ⁓ you see that you have ⁓ key advantage that you are based here in Europe ⁓ to the Chinese or other competitors that ⁓ come to Europe as well? speaker-1: be honest, of all, this market is so big and so large that there enough business for everyone. I think behind the scene, people publicly there is a lot of competition. Of course, we are competitors, but we are also friends. We are having a very close relationship to our competitors being from the US, China, doesn't matter. Because if you are deeply in this business, you know that... how large is the market and there is not one player in all business. So we need multiple solutions. I think over time we will have the different segments of the markets. Sometimes you need to have premium solutions because the requirements of the use case are so high. Sometimes you have to go with maybe a bit best in cost category solutions. So we start to ⁓ already some ⁓ of the patterns. that different providers from different regions of the world are going to different categories. And I think there is just normal what is happening on the market. All people are looking to the European market. It is a big market as we are looking to the other markets, for example the US. And the fact that, for example, the Chinese colleagues are quite interested in Europe for me is making sense. It's very logical. Because the driver shortage and the driver cost... and the business case is working much better here in my beliefs. And they have the advantage of their local market that they can develop, they can test, they can deploy, and then come with the perfect solutions here. And I think, again, if you look to the logistics sector, we are still talking about a handful. And for a worldwide market, I think we are going to see a market segmentation categorization. speaker-0: And maybe a bit of ⁓ a question. So what is the one thing about running an AV company in Europe? That you wish more people in the industry or also in the media, press, politics, can actually better? speaker-1: Yeah, I mean, for Europe, think would be important that the whole ecosystem from automotive industry and also on the political side to understand the chance that we have to transform because the whole transportation companies, industries, including it's a big piece for the European economy. So and I think the big shift we see is happening thanks to the automation. We have to realize that. And I think we have been sleeping for some years, but I have the feeling that now we are waking up. And that's always a good sign because we are very ⁓ here in Europe by culture. But when we are waking up, then we are very serious. then ⁓ have to see a bit also what our OEMs are doing, but we see already the signs, the big OEMs, they are already planning to go into level four. That's great. ⁓ And we see also different jurisdictions that we have in Europe. They have already prepared their legal framework like Germany, like France, like Switzerland and so on. have just seen this morning, for example, even Portugal, they have started to allow the autonomous driving testing. These are the perfect signs that we need in Europe to open up the market, because very often we compare ourselves with the US market. But even if you look at the US market, we are still not talking about ⁓ single state allowing autonomous driving. So We don't have to get always this picture that we are late and so on. We have just different ways of doing things, but we have to be fast. We have to be fast. We have to make sure that we don't create too many interdependencies you know how the political scene will And it's also an asset that have to keep. We have excellent engineers. I think this is a big opportunity for whole industry in automotive to transform and to develop new competencies and to keep ⁓ this great industries that we have, we have obviously some challenges for electrification, but now I see this automation as a big opportunity. ⁓ this is ⁓ direction that I wish that our ⁓ goes, that our politician goes. But yeah, sometimes we are a bit slow, but when we wake up, we are serious. ⁓ speaker-0: Perfect. So maybe to wrap everything up, what is the one thing that you want our ⁓ to remember after this podcast in regards to LOXO? speaker-1: you know, ⁓ I would love to communicate here is level of performance that we have, you know, because very often in the people who are technically and deeply involved, they see a strong correlation ⁓ of, for example, size and performance. This is not the way that we have as a ⁓ point view. So let's call it the Swiss way. So we are, we see ourselves as extremely efficient ⁓ are going to communicate quite soon the main KPIs of our company in a transparent way. And there are some other companies also communicating the same things. And I would like the listeners to go check that and compare it. And if you have the opportunity to take a ride with us in Bern, ⁓ a month we do ⁓ open door day. It's for free, you can register, you can come, you can see the whole logistic process, you can have a live view of the car, how everything is working. this is ⁓ impression that I would like to give, because think we are quite ahead of what is prepared so in Europe. And yeah, I would like to use this platform to communicate this message to invite people, especially Europeans, have a look to what is happening in Europe, and you can be proud of. Yeah. speaker-0: They're really good closing words. So I will definitely check out your KPIs and also compare it in my newsletter for sure. Okay. And also ⁓ visit you and Bern think it's ⁓ great offer from your end. So we're now at the end of the podcast. So thanks Amin for taking the time today here at the Automated Mobility Summit in Zurich. We now have the whole day ⁓ with great and keynotes and discussions ahead of us. But yeah, thank so much for joining us. speaker-1: Thanks for having me. It was a pleasure.