Transcript: Click a Button Get a Car

Executive Summary

In this episode of The Road to Autonomy podcast, Grayson Brulte sits down with Justin Spratt from Vay where they discussed the details the company’s unique, safety-first strategy for teledriving.

Justin explained how Vay built its automotive safety case before scaling, how its proprietary VayNet with redundant networks minimizes latency, and how its flexible business model—offering both direct-to-consumer services and a B2B licensing platform is poised to unlock massive economic efficiencies across car sharing, logistics, and autonomous vehicle remote support.

Key The Road to Autonomy Episode Questions Answered

Key The Road to Autonomy Topics & Timestamps

[00:00] Vay’s “Automotive Engineering First” Approach to Teledriving 

Justinn Spratt explains that Vay’s core strategy is fundamentally different from many competitors. Instead of prioritizing software, Vay began by building a robust automotive safety case with its own hardware and safety controller. This deep focus on automotive-grade safety from day one is what Justin calls a key part of their “secret sauce,” providing a solid foundation before scaling their operations and software.

[01:15] Why the Founders Prioritized the Difficult Automotive Safety Case 

The decision to focus on safety first was driven by the founders’ extensive backgrounds in the automotive industry. They recognized that the software component of teledriving is relatively easy, but the true challenge—and the key market differentiator—is proving it’s safe to operate in complex urban settings without a driver physically in the car. This difficult and expensive path was validated when German authorities granted Vay a permit to operate on public roads, a testament to their high safety standards.

[02:40] How Vay Solves Latency with Redundant Networks 

Latency is a critical safety element in teledriving, and Vay tackles it with a multi-layered approach. Their proprietary system, VayNet, uses advanced compression and channel multiplexing to ensure a smooth data stream. To prevent connection loss, each vehicle is equipped with four modems capable of running on four different cellular networks simultaneously, making a total failure “incredibly improbable”. As a final failsafe, if the connection is ever lost, the vehicle’s onboard software can bring it to a safe stop on its own.

[03:55] The B2C Model: “Click a Button, Get a Car” 

Vay’s direct-to-consumer service offers the convenience of ride-hailing at a fraction of the cost. The model is simple: a user requests a car, it’s remotely driven to their location, they drive themselves, and when they’re done, a teledriver makes the car “magically go away”. By having the user drive the main leg of the journey, Vay can offer the service for approximately 50% less than ride-hailing while maintaining a profitable business model with over 20% headroom.

[04:55] The B2B Model: A Flexible SaaS Platform for Partners 

For business partners, Vay offers a powerful way to reduce operational costs and improve efficiency. They can provide a full turnkey, software-as-a-service (SaaS) solution for teledriving, which can save partners up to 40% on their driving-related expenses. Recognizing that not all partners have the same needs, Vay has also “decomposed” their technology stack, allowing companies to license individual components like the networking stack or the human driving element as needed.

[06:45] How Customer Demand Pulled Vay into the B2B Market 

The development of Vay’s B2B licensing model was not part of the original roadmap; it was a direct result of “customer pull”. After focusing on their consumer pitch, Vay received a massive amount of inbound interest from a wide range of industries, including car sharing, logistics, trucking, and even autonomous vehicle companies. This overwhelming demand made it clear there was a significant market opportunity for a flexible, licensable teledriving platform.

[08:18] Scaling Operations with Multi-Vehicle Teledrivers 

Vay’s plan for scaling involves creating centralized operations where teledrivers can operate a wide variety of vehicles for multiple partners from a single infrastructure. A driver with the proper licenses could seamlessly switch between driving a passenger car for a rental service and a delivery van for a logistics company. This shared-resource model is key to maximizing efficiency and driving down costs for Vay’s B2B customers.

[10:45] The Future Use Cases for Teledriving 

Justin Spratt envisions teledriving becoming a multi-billion dollar category that is integral to nearly every form of future mobility. Key applications include assisting autonomous vehicles with edge cases they can’t handle and solving the “first and last mile” problem for autonomous trucking by remotely navigating trucks off highways and into distribution depots. This human-in-the-loop capability makes it quicker and more economical to scale than training an AI for every possible scenario.

Full Episode Transcript

Grayson Brulte: Justin, Vay, you’re, growing. You’re building. I was in your car last year at CES and it performed extremely well. How is Vey approaching teledriving? 

Justin Spratt: Thanks for having us on. We, we’re doing well. We’ve, we’ve, you know, hitting into our 5, 000th trip in Vegas. We’ve just decided to scale out our car. Our fleet there to about 150 vehicles by the end of Q2. So very, very interesting times, very excited. The way we approach tele driving, unlike some other companies is we started with the automotive engineering piece first. So fundamentally we have made safety core to the business. And that means we have built an automotive safety case, which then also has hardware and automotive. Safety controller. , and that really is a lot of our secret sauce. How do we deal with the different engagements that a tele driver would have, all the way through, you know, from the latency stack down to minimum risk maneuvers. So that’s it. And then we’ve wrapped and kind of augmented the Silicon Valley software, talents around that. And, and, you know, 5 to 6 years later, here we are, and it seems to be, you know, that why that was a bit of a longer path, and more expensive. It turns out that that seems to be the path that was the right one, given our market position.

Grayson Brulte: Why was the decision made early on to focus on the automotive safety case first? 

Justin Spratt: Of the three founders, two of the guys had worked in the automotive space, one AB automotive and then Fabrizio, the CTO, had worked across all of the German automotives. And they realized very early on that to make this thing work with software is actually quite easy. , but then the ability to put people in cars without safety, driver safety, operating in urban settings. Actually, the hard part is the safety part. And so very early on, they realized that that would be probably the key differentiator. And so they spent a lot of time, they essentially slowed down, spent a lot of time on that and then accelerated once they’ve got that done. And that if you know anything about the automotive space, that safety concept is, it sounds like a fairly easy thing to do if you’re a non automotive person, but it’s incredibly difficult, huge amount of money. Lots of testing and validation. Then, obviously, we had the validation from the German, roads authority, across three different organizations, and we got a permit to drive without safety driver or safety operator in and public streets. And, you know, Germany arguably has the high safety standards, , road safety standards in the world. And so that was, , testament to the strategic decision.

Grayson Brulte: It was a strategic decision to say the least, as you’ve experienced staying on the safety is where we get to business. Since you’re teledriving, one of the critical safety elements is latency. How do you limit the latency to be as little as possible? 

Justin Spratt: Yes. So there’s a bunch of elements. And you probably need someone smarter than me to go into detail, but essentially it’s around how you do your compression, how you channel bond and slice, and multiplex the different radio frequency channels . and then also about making sure you, you. Don’t have a single single point of failure. So we have redundant networks. Our particular hardware is for modems. And so, in theory, we can run four different networks. And so to lose contact with that car, you would need four networks to fail, which is, you know, incredibly improbable. And then we’ve obviously built. Kind of some safety software in there. We also have our own, you know, AI and autonomy team. And so if we do lose contact with a car, we can bring that car to a safe stop. But so there really is a lot of safety layers there, but fundamentally, it’s, it’s, it’s about that network stack. And we have, I would argue, and it’s been validated by the MNOs, the mobile network operators we’ve chatted to. We have probably the best network engineering team when it comes to this kind of, service in the world. And these guys have spent a huge amount of time building out what we call VayNet. And, yeah, that’s, that’s a lot of our secret sauce. It’s hard.

Grayson Brulte: It’s very hard to say the least. You have VayNet, you have the automotive safety, the next piece of the ingredient. What’s the business model? 

Justin Spratt: It’s quite cool. So I’ll tell you a little bit of the B to C pitch, which is what we raised our Series B on. It’s kind of like, click the button, get a car, you know? And so if we bring you the car, tele driven, you drive yourself. And then we make it magically go away. We’re going to charge you around about 50 percent less than ridehail. So, on that basis, we still have a 20 plus percent headroom on, on the business model there . so it’s kind of super interesting that if you can do the trade where the person drives the car, you end up having the most affordable mobility door to door mobility service in the world. And in a price sensitive market. That’s really interesting for people as we’re seeing in Las Vegas, you know, with our current fleet and then driving a ton and 50. Now, if you have that as your base cost structure, then when you want to sell it to our B2B partners, we find out that, if we put the tele driving, the human component on our platform, there is actually an incredible amount of scale there. And so if we tell you drive on behalf of our partners, you can save up to 40% on the driving component cost. And you know that driving component cost is really, difficult. The supply of drivers is declining. It’s a super tough job. And so, you know, if you can get that thing not only working, , in a high functioning way and save up to 40%, it’s super interesting. And then the way in which we offer that to our B2B customers is quite simple. We have a turnkey solution, so kind of like software as a service for tele driving. Basically, we can do everything for you, you know, the drivers, the network across the software, the telestations, the network connectivity, all of that, and then you pay for what you use. But we’ve also then kind of decomposed each one of these product layers and we can, we’ve essentially productized all of that across five verticals. And so you can actually license or rent a key component. So if you only needed our networking stack, we could, we can, do a deal on that basis. If you just wanted the human tele driving element and software elsewhere, we could also do that. In fact, we have a, an AV partner whereby we’ve just done a deal where they’re going to take the software components and the network components only. We found that the B to B partners, the commercial partners, enterprise partners want that flexibility because, I mean, if you think about this, it’s uncertain how this thing plays out for them. If they’re going to have massive scale, they will build their own operations teams, no doubt. But do you want to go and deploy multi million dollar infrastructure without knowing what and how big this is? I think that’s still the question, and that’s where we offer our platform, which gives them that cost saving.

Grayson Brulte: Licensing is going to be the most profitable aspect of autonomy. You have individuals that are developing the full driving stack, that license to become very profitable. You have companies like Vay, that are developing the tele driving stack, which become very profitable. Was licensing always on the roadmap from day one? Or is that a recent development since you’ve joined the company? 

Justin Spratt: No, it wasn’t initially, it’s a recent development and , it’s got nothing to do with me. It’s actually customer pull. So what we found, in fact, we weren’t going to launch B to B. We had this B to C pitch, a and that was the business model. And then we had a ton of inbound from car share, car rental, logistics, trucking, and then autonomous vehicles and governments and all this inbound. And we’re like, wow, geez, these are lots of different, go to markets. It’s actually, there’s a lot going on here. , and so what we’ve done, my, my BD team has spent the last 12 months engaging these partners to try and find out what. Is the strongest signal. And it turns out that the common thread through all of those is the licensing piece. And so at some point as teledriving becomes more prolific and there is this kind of adoption, the licensing piece will get bigger. But I think fundamentally, we still need to offer our partners the on ramp of being able to give this so called turnkey solution . it’s incredibly interesting, you know, the way I think this thing plays out. I mean, you know, we might talk about that even later, but is that to your point? It will be licensing. You’ll find these companies start getting incredible scale on their own tele driving. And so they will want to have their own operation centers, similar to like they have their own driving, teams, and then eventually it’ll be a software kind of tele processing software cloud based service that we that we offer, but still a way to go until then.

Grayson Brulte: How many teledrivers you need per vehicle? And could those teledrivers drive hypothetically different vehicles from different partners? Are you going to build a teledriving warehouse where individuals are driving various different vehicles? Or how is that going to work as you scale and as your partner scale? 

Justin Spratt: Yeah, it’s a great question. We’re going to do exactly that. We’ll have telle drivers that have multiple driving types of driving license and they will be able to drive multiple, myriad selection of vehicles for myriad partners. And, we will put it on, you know, as much as possible, a single single infrastructure. We might do separate instances. Okay. So if you have a luxury carmaker and they want a white glove service, then we can, you know, build out a separate instance and allow them to have their own customer script and engagement model. But fundamentally to drive the economics on this thing, you want to try and put it on one platform. And so that’s, that’s kind of what we are doing. And we’ve already inked deals and signed deals, that let us do that. And, and again, obviously the economics there passes through to the, to the customer.

Grayson Brulte: What are the economics of teledriving? 

Justin Spratt: The stack, you know, as I said, the plain vanilla version, which is freeing up the empty miles from, say, someone sitting in the car, is about 40%. We, that’s on the B to B basis on the B to C basis. If we run our own fleet, then we can make, you know, I won’t disclose it fully, but we, we make good margin fundraisable margin at 50 percent lower than ri So, and again, it depends on the city and the complexity of the operating domain. But the rule of thumb is we make, we have a profitable business at half, you know, half the price of teledriving. And the reason for that, as you and your listeners will know is that we are kind of bringing those empty miles close to zero. And that’s the efficiency that we get into the system. That’s before we have our augmented kind of software enhancements and some of our AI.

Grayson Brulte: Your monetizing empty miles, which raises the other question. What’s the use case for Vay is it a limited use case that a wide use case or what type of customers do you see using bait today and tomorrow? 

Justin Spratt: like I was saying, we’ve had inbounds from, you know, luxury car segments and premium car segment. We’ve had from car share and car rental across the globe. We’ve had, , light commercial vehicle and middle mile logistics, supply chain logistics, we’ve had, shipping ports, airports, off roads, and then long haul trucking, and we’ve literally had inbounds on all of that. And so, I think fast forward, You know, three to five years, there’ll be a billion dollar, a multi billion dollar tele driving category, and it will probably span across all of those use cases. Because even if autonomous vehicle autonomy works in a lot of those use cases, it doesn’t make sense to train models to do some of those use cases. It’s just a lot more economical to have humans do it a lot quicker, a lot easier to scale. So we think it’s across all of that. I personally think that the, you know, the car share use case, , is interesting, you know, a la Zipcar, is very, very interesting. And we, we see that as being able to unlock new total addressable markets for, for that, those kinds of companies, because all of a sudden it’s click a button, get a car. There’s none of that friction, and then, but I think that the autonomous vehicle category is very, very interesting for us. And we having a lot of really productive engagements there . around assisted autonomy. So basically where, where you can’t train a model because it’s impossible or too expensive, then the tele driving can obviously fit there. And then, of course, if the car gets stuck, there is a system failure. You know, tele ops or tele assist doesn’t work. You know, you need to actually be able to drive the car. And so we’re starting to see that the tele driving or the remote driving capability is core to the success of of autonomous vehicles being able to scale up. So those are two very, very interesting categories. And then trucking is also is also very interesting. Because the trucking parts on, and you’ll know this better than me, the trucking part of autonomy. Is is pretty much solved on freeways. They can do that incredibly well. The difficult, the difficult bit is the ingress and egress off the freeways to the distribution, depot and that is complicated. Again, you’re getting into urban urban settings and it turns out that, a focused human being is actually a pretty good sensing and perception engine. And so, you know, we just essentially basically, , teleport in, take control of the truck without it even stopping. And we’re able to drive it to the depot and then from depot back on the freeway, then it automatically, it seamlessly switches to autonomy. And so we’ve proven all of that out. So I think that’ll be a huge category as well.

Grayson Brulte: You’re enabling autonomy companies in the early days of autonomy. If you want to use the term that’s used all the time, Level2+ or Level 3 to scale. Overall, how do you see the telemarket evolving and changing over the next decade? 

Justin Spratt: Yeah, you know, like I was alluding to earlier, I think it’s a multi billion dollar category. I think, you know, everywhere, like, like seaports, they’re just like unmitigated, logistical disasters, and you have like 20 to 100, 000 people going into them and people are parked everywhere. It’s, you know, if you could have some kind of tele driving system there. You could move people way more efficiently there. And it’s a very concentrated area. It’s it’s it’s it’s more private grounds. That’s super interesting airports and baggage transfer, can be done there too. It’s just a lot cheaper. And so I think this, you know, car share, car rental, it probably needs a lot of work. Car rental is not an easy one to do because you need quite a few cars to kind of affect the shape of that market. But if you imagine car rental, all of a sudden you don’t need physical locations. You literally have dynamic supply wherever you need it in a city between airports and city like that. That’s a game changer. It’s just going to take a little bit longer. And then trucking and then, and then, uh, the, the autonomous vehicle space. So I think it’s a multi billion dollar category that say in five years, and it’s going to be integral to pretty much every form of future mobility, you know, from autonomy, even through to kind of electric buses in suburban areas. Through the ports and what have you. It’s incredibly interesting space and there’s just so much going on here. It’s hard to kind of give it a narrow definition, frankly.

Grayson Brulte: The bottom line is teledriving unlocks economic efficiencies. And Justin, we’ll have you back on late in October to have a full conversation about Vay and where you’re going. The future is bright. The future is autonomous. The future is Vay. Justin, thank you so much for coming on The Road to Autonomy Insights today.

Justin Spratt: Thank you very, very much for having us. Thank you.