 From the SiliconANGLE Media office in Boston, Massachusetts, it's theCUBE. Now, here's your host, Dave Vellante. Hello everyone, and welcome to this CUBE conversation. You know, the auto industry was a, if not the, dominant force in the 20th century economy. And clearly, you see it in the headlines today. I mean, all you got to do is look at Tesla. The stock is absolutely on fire. Tesla's market value is actually greater than that of Ford and GM combined, even though its revenues are about one-twelfth of those two combined. The macro discussion today is really heating up around ESG, which stands for environmental social governance. So electric vehicles are really picking up momentum. And you know, maybe that's the tailwind for Tesla, but consumers are pragmatic. The electric is still more expensive than internal combustion-powered vehicles. So we'll see how that plays out. One of the things we talk about a lot on theCUBE is the software content in automobiles. In many ways, these vehicles are code on wheels. So that's part of the hype factor too. But you know, I've always argued that the incumbent automakers are actually in a pretty reasonable position to compete. While autonomous vehicles, they may disrupt the incumbents. And even though right now Silicon Valley is ahead of Detroit and Japan and Germany and Korea, there's an ecosystem that is evolving to support traditional automakers. Now, one of those players is tactile mobility. The vast majority of data created around autonomous vehicles today is visual-based with LiDAR as a key enabler. But a human driver, you think about it, they don't just rely on sight. They're able to feel the road, the bumps, the curves and the impacts of things like weather. In fact, it's estimated that more than 20% of vehicle crashes in the US each year are weather-related. In intelligent cars, they really still can't predict road conditions ahead. Tactile offers software that uses sensors that already live in the vehicles to predict and feel road conditions like black ice and potholes to improve safety. And with me to talk about these trends and his company is Amit Nisenbaum, who's the CEO of Tactile Mobility, Amit. Thanks so much for coming on theCUBE. Thank you, Dave, very much for having me. Yeah, so really it was a great opportunity when I heard you were in town, invited you out and really appreciate you coming out to our Mallboro Studios. But let me start with why your founders launched Tactile Mobility. Well, Dave, it's a very interesting story, I think, for our company as well, for other entrepreneurs to learn from it, because actually the company has been around for about eight years and it all started from a conundrum from a question that was posed to our founder, Bos Mizrachi, which was about how do you take a vehicle from point A to point B at a set speed with minimum gas consumption, using only software and data coming off the vehicle sensors that are run of the mill sensors? And that question started this whole company. He believed that it's only an optimization question, meaning all of the data is out there. The meaning data about the conditions of the road, the grades, the curvatures, the conditions and the health of the vehicle, meaning engine efficiency, tire health, et cetera, et cetera. And what he found out was that actually neither this nor that has existed. So it was way more complicated than an optimization question. It's about how do you generate that data about the vehicle and the road? And he launched the company in order to go after those two data sets. He was able to solve that or to address that question and to take a vehicle and to show that you can take a vehicle from point A to point B at a set speed while minimizing fuel consumption up to 10%. By the time that he has done that, gas prices dropped and the question was what's next? And fortunately enough, the industry and the hype around autonomous vehicles has come around and that has been the next frontier for our company. And that's what we've been focusing on since then, but not only on that, but also on other aspects which I'll be happy to speak about. Yeah, this is an awesome story of a pivot. I mean, you see this all the time with startups. It's going to survive until you can thrive. And then something happens. It's a tailwind, great technology that the visionary can see how to reapply it and a little bit of luck involved. Maybe, okay. Stamina. Stamina, right, right. You have to have a strong heart to be in stomach to be a startup. Okay, and then you joined just a couple of years ago. That's correct. What attracted you to tactile? Well, I've been in this industry actually in the cross section of the two industries of automotive and energy for about 12 years now, starting from a company called Better Place that you might have heard of. I was one of the first hand employees there. And those two industries have been near and dear to my heart ever since. I like big questions. I like big challenges. I like big plays that have the potential to make a real difference. So the fact that the tactile mobility at the time it was called Mobywise was in this industry was a big plus. But also the fact that the offering is not really the vanilla flavor offering. Everybody's doing LiDAR and radar and cameras. All of a sudden there is someone else that is saying, women, there is that neglected segment that additional set of senses, the additional, the sense of tactility that all of us are using when we're driving and computers will need that as well. How about that? This is something that nobody pays attention to. And that really caught my attention. So I kind of hinted at this in my little narrative up front. The hype was all around autonomous but let's face it, level five autonomous, it's at least 2030, maybe further. But everybody drives some form of autonomous vehicle today if you purchase a new vehicle and that's really the space that you play in. So what are the big trends that you see and what's the problem that you're solving? Yeah, so first of all, you're absolutely right. When people speak about autonomous vehicle, they imagine to themselves a car, a vehicle with big red button and that's it. That's what is called level five. However, there are four levels below that that lead to that and today, most of the vehicles leaving the assembly line are either level two or level three. That's why we're also saying that we're in the business of smart and autonomous vehicles and the challenges there if you're looking at the vehicles themselves are challenges of how do we make those vehicles both safer as well as more enjoyable to ride and the ability to address both of those together is actually not as simple as one might think. So that's what we're focusing on and that's the trend, the trend of no compromises that you go both for safety as well as a user experience. That's on the vehicle side. Having said that, being a data company that has a proprietary software stack that allows it to generate that data, the tactile data, the data about the dynamic between the vehicle and the road allows us also to take that data to the cloud and in the cloud to split that dynamic into two separate models. One we model independently the vehicle, the vehicle health and the other one is we're turning each one of the vehicles to become like a probe that fills the road conditions and maps, the location of bumps, cracks, oil spills, black ice, et cetera, et cetera. And by that we are able to crowd source and create the data and create new layer and new of the map road conditions there. Going back to the question that was posed about how do you take that vehicle from point A to point B, a minimum fuel. Here you go, we have those two types of data and now we can use it in other verticals as well. Well, that's very interesting. So a lot of people say, oh autonomous vehicles it's all about real time, you can't do anything in the cloud and you actually, you're refuting that because you're building essentially a map of what's happening on the roads, whether it's a pothole or a bump or a curve, et cetera. That's correct. And so essentially you're doing that in the cloud, modeling that in the cloud and then what, bringing it down in real time, right? Yeah, so first of all the first use case is indeed to bring it back to the vehicles and so the vehicle and the vehicles around it will know what's ahead of them. Use cases there are about preconditioning vehicle systems. For instance, you're approaching a pothole, probably you want you, meaning the vehicle would like to tune the suspension to become harder or softer. You're approaching black eyes, probably you want you, the vehicle would like to slow down. So that's one use case. But there are other use cases, other use cases around for instance, road authorities and municipalities. We do have customers around the globe, road authorities and municipalities that are subscribed to our data services that allow them, road condition data services that allow them to better plan maintenance as well as dispatch crews to locations of hazards in real time. Yeah, so I remember when I was a kid, we had a CB. That's how you communicated what was ahead. Hey, watch out, there's a pothole up ahead. Right, analogy. Yeah, and now we're doing that. And now does that essentially require some kind of peer-to-peer network? So we're agnostic of the technology. We're the data layer behind all of that. These days everything or most of the use cases are still running on a vehicle to cloud to vehicle or to anybody else. But there are companies that are working on vehicle to vehicle. So you mentioned the stack. What does your stack look like? Can you describe that a little bit? Two parts. One is embedded software that sits on one of the vehicle computers, one of the ECUs. And the other one is the cloud component. The component, the embedded software that sits on one of the vehicle ECUs, usually either the gateway or one of the vehicle dynamics ECUs or maybe ADAS say ECU, et cetera, takes in real-time amounts of data from multiple existing non-visual sensors such as wheel speed from all four wheels, wheel angle, position of the gas paddle, torque of the brake paddle and much, much more ingest all of that, create a unified signal that describes in real-time the dynamic between the vehicle and the road. That signal is very, very noisy. So we apply signal processing methodologies to clean it. And then we apply on top of it, algorithms and AI and all of that in real-time in order to derive insights about the vehicle road dynamics. You probably ask yourself, give me a concrete example of something like that because it's kind of amorphous. The killer up to these days with OEMs, the vehicle manufacturers, is what is called available grip level. It's basically a signal to the vehicle computer about how drastically can the vehicle accelerate, decelerate or change direction, all different types of acceleration before it will start to skid. Think about it as the performance envelope of the vehicle. Nobody but us can model this using software only in any condition and this type of data has multiple use cases in the vehicle. Happy to tell you more about those. Question is if do you have time? We do, but I want to make a point, the software only, the thing, if I understand it correctly, the OEM doesn't have to change any hardware. You're using the existing sensors of the vehicle of which there are certainly dozens, if not hundreds, to actually take advantage of this, right? You don't have to do any kind of hardware changes, is that correct? We're data and data analytics and AI companies. Yeah, so if you want to add some color and double click on some examples, that would be great. Sure, so going back to the available grip level type of data of insight, I call it, think about adaptive cruise control, the function that allows a vehicle to drive at a set speed, however, to avoid colliding into the front vehicle. So today, it seems like all of the data is there for ACC adaptive cruise control to be effective. You know the distance from the vehicle, probably using a radar. You know the relative velocity between the two vehicles, so you have all of the information. However, you don't know you, again, the vehicle computer, how hard the vehicle can break, given how slippery the road is, given how healthy or worn out the tires are, et cetera, et cetera. That means that the vehicle computer needs to err on the safe side and keep the large distance in order to allow safe braking. What's wrong with that? Going back to the question about the trend before. First of all, it's not natural to the driver. We keep a certain distance for a certain reason, and when the distance is too large, it just doesn't feel natural to us. That's one thing. However, on the other side, it's also not safe. How is that? You keep too large of a distance, someone at the end will cut your end. And ironically, you kept a large distance to stay safe, all of a sudden, your war's off. Being able to allow the vehicle to know really what is the tight distance, safe distance, to stay from the vehicle allows that vehicle to be more enjoyable to ride as well as safe. So take that example, because today, I can sort of personalize that adaptive cruise control and say, okay, I want one bar, two bar, three bar. Yes, that's it. But that's it. That's it. I sometimes say, well, is three bar right? Is two bar right? You're right. You go, well, that's too far. I think I'll cut it down to two bar or one bar. You're saying with your software, the system is intelligent enough to optimize that, keep me safe. That's correct. But also keep me having comfortable driving. Absolutely true. Actually, those three bars is kind of a psychological exercise, right? Because the shortest bar is that large distance. When they tell you two bars or three bars, it's kind of like, do you want to keep a large, very large, or extra large distance, right? Because they will never allow you to keep shorter distance, shorter than what is really, really the bare minimum in order to break at the worst case scenario. Even if it's safe. Even if it's safe. And that's really where your software comes in. Absolutely. Now, Porsche is an investor in the company. That's correct. Presumably, it's a customer, right? No, they actually said publicly that they're a customer as well. Okay, great. Let's talk about how customers are using this and what the adoption cycle looks like. And maybe you could give us some examples of how it's being applied. So customers, you mean OEMs, car manufacturers. So the way that they use it, I just described it now. The adoption cycle, in this industry, unfortunately, cycles are long. We work years to create relationships with the car manufacturers, to allow them to learn about our capabilities, to validate the integrity of our software. They also most commonly run RFPs or RFQs in order to choose the right technology. And I'm glad to say that we're winning again and again and again. And then there is the integration cycle, which by itself is a few years in length. So the cycle altogether is long. However, we found that our approach is quite effective and the approach, not necessarily the technology, yes, but also the way that we approach those OEMs. We are quite, if I may say, humble. We know that we're not the car engineers, the typical car engineers. We actually know very little about cars. What we know, we know data very well and we know AI very well. And when we come to them, we say, we're not trying to replace your engineers. We're not trying to do what you do. We're trying to tackle the same problems that you weren't able to tackle before from a very different angle. And that works very well. So you talk about the integration cycle of a couple or maybe even longer. How long is the design cycle for these things? Is it also years or? So the design cycle from our perspective is much, much more agile. Actually, we are working in agile framework in terms of the development of the software itself, but you're asking about the design much faster. But when I said a few years, a couple of years, I meant a per OEM to design together to allow them to feel that we're designing, meaning customizing the software to their needs as well as implementing it. That's the level. But what they get is a competitive advantage. So Porsche as a leader, obviously, and an early adopter is going to be able to now commercialize this technology. Of course, it'll be embedded, but that'll be a feature that the car sales person will highlight and maybe they market it, maybe they don't. But that gives them a competitive differentiation. So are you seeing that other OEMs are starting to really get this and sort of leaning in or what's your experience? Yes, it's the typical technology adoption curve. There are the early adopters and there are the mainstream and the late adopters. I'm glad to say that these days we're not only working with the early adopters, but also more with the mainstream. I encourage you to stay tuned. I believe that in the coming month or two we'll have a big announcement about another major OEM that has chosen us commercially for mass production. And we are in quite advanced stages with OEMs both in Europe and North America, starting also to spin out to Asia. And is the business model, is it a subscription model? Is it a one-time payment from the OEM? How's it work? That's another thing that made me excited about the company going back to your question from before. It's quite diverse, I would say. For the OEMs, the software that we embed in their vehicle, it's software licensing. However, the data that we generate and then upload to the cloud and repurpose it with the OEMs themselves, but also, as I said before, road authorities, municipalities, fleet managers, insurance companies. I didn't have a chance to touch on all of the verticals. That's a subscription model. So the two models are working together. It's actually quite an attractive value proposition for us and for our investors. So there's software license and there's data as a service. Absolutely. And there's also adjacent industries that you can go after. You just mentioned a couple. So when you think about the total available market, which obviously any CEO is going to do and TAM expansion is part of your job, but so what's that vision? When does that look like? So in terms of the size itself, it's measured in the trillions. It's very, very big. In terms of the different verticals, the ones that I tapped on are the first ones, but even within those these days, we're really, really trying to stay razor focused on the OEMs and road authorities and municipalities. We have fleets and fleet managers that are coming to us when we request for the data that we call vehicle DNA. That's the data about the vehicle health, et cetera. And that's the third vertical that we're starting to address these days, but we're only 25 people growing to 40. We're trying to be very, very agile. That's from one end and from the other end, now that we showed our value to the car manufacturers, we're going for the E-Force multipliers, meaning partnerships with the channels, with T-1s, the suppliers to the OEMs themselves. And let's say you've been around eight years, you've been there two years, right? And then I think you did a raise of roughly what, nine million today? We in October, yes, October 2019, we announced the latest round of $9 million from Porsche as well as some other investors, yes. Yeah, great, okay, so I mean, not a ton of money, but you guys are small, and so a little bit more on the company. You said 20, going to 40, you're well-capitalized, but today you see people raising $250 million. Yes, that's right. What do you sense is your capital needs? I mean, you're obviously actively raising money and doing as what the CEO does, but can you share with us kind of your milestones for the next 12, 18 months? First of all, we were fortunate and fortune has something to do with it. I think that being disciplined is another thing, to have revenue already. So our capital needs were still not profitable and we're growing fast, so we need to raise in order to support that growth, but we're quite diligent about that. Also, true, companies have raised tens and hundreds of millions of dollars. First of all, not all companies in this industry are created equal. We're not a hardware company. We're software and data. We're also not trying to do a fully integrated offering like, let's say, Zooks or something like that, which requires way, way more money. And actually, I'm quite glad that we're raising as we need, but not more than that, because what you raise, you need to return tenfold. So we have enough in order to support the growth of the company in years to come. Well, the OEM model is very sales-efficient as well, right? I like them in the software companies today are hiring people to inside sales, outside sales, enterprise sales, and so there's a different business. Well, first of all, congratulations. A really interesting story. Really appreciate you coming out to our studios here in Marlboro and sharing your story. Best of luck to you. Thank you very much, Dave. It's been a pleasure coming here and I'm glad that you invited me. Great, and thank you everybody for watching. This is Dave Vellante with theCUBE. We'll see you next time.