Tyler Suiters:
Hey everybody, I'm Tyler Suiters with the Consumer Technology Association. We are the owners and producers of CES, the most influential tech event in the world. We are getting you CES ready. Our all digital show runs January 11th through the 14th 2021. We're re-imagining CES. Of course we're supporting what makes CES unique, but we're also enabling you to connect with new global audiences. Exhibitors can launch the latest products and you can connect with global brands, hot startups, and one another. You can get business done at CES. This will be a new experience. You'll get a front row seat to discover and see the latest technologies. You'll find captivating imagery, dynamic presentations, engaging content, and the key categories that drive it all. I think you know the well by now, drones, 5g, smart cities, digital health, resilience, applications of AI and vehicle technologies.
Tyler Suiters:
Those last two categories, AI and vehicle tech, they are our focus for today's episode. Today a conversation with SafeRide Technologies out of Israel. Yes, this is about protecting our cars, our vehicle technologies, as our cars get more connected, able to do more on their own, but SafeRide has a specific focus on vehicle health management. So a deep dive conversation today into what AI and deep learning mean for the increasingly connected automotive sector. Today, a conversation with SafeRide Technologies on this edition of CES tech talk.
Tyler Suiters:
Joining us today from Tel Aviv, Israel is Hilik Stein. He is the CTO and co-founder of SafeRide Technologies. Hilik, I say, good morning here, it's good evening for you. Hope everything is going well in Tel Aviv and welcome to the podcast.
Hilik Stein:
Thank you. Nice to be here.
Tyler Suiters:
Let's start at the beginning. An overview of what SafeRide offers. The name itself indicates vehicle technology and the future of vehicle technology, but you operate in a specific niche and it's something that is not just a technology that is on the way, but a technology that is already here in terms of connected cars and the need to protect them.
Hilik Stein:
SafeRide is dealing with vehicle health management. We're focused around connected vehicle and in the future also autonomous vehicles and providing solutions to create health management solutions for these vehicles.
Tyler Suiters:
A driving need. So I don't mean to play on words there, but the idea that as we move towards self-driving vehicles, the need for vehicle health and cybersecurity grows ever stronger. But in truth, we're already there. I mean it's 2021, there are so many connected cars or elements of conductivity and the cars were already driving. That vehicle health and cybersecurity are concerned today, not just tomorrow, right?
Hilik Stein:
Absolutely. I think the first time it was actually proven publicly was back in 2014 with the famous G pack in what [inaudible 00:03:40] black hat conference, some academic paper actually showed how a vehicle could be taken control of remotely. Pretty scary stuff.
Tyler Suiters:
Was that something of an impetus for you to co-found SafeRide? You were established in 2016, so not that long after that demonstration of hackability if you will. I assume that had a pretty strong effect on you and your team.
Hilik Stein:
Yeah, so actually SafeRide started back in 2015 when my co-founder long friend, longtime colleague Yossi Vardi and I were sitting together at a Starbucks in Silicon Valley where I was living at the time. Suddenly, so all of the autonomous pilot vehicles by Google and others driving around us, we've noticed the huge number of cameras, the sensors on control of mountains on this vehicle and started wondering what else could be done with combining AI based insight with the enormous amount of newly available data on this vehicle. We were then established early 2016 with headquarters, including R and D in Tel-Aviv and myself running the office in the Silicon Valley until I moved back to Israel in 2018. Since then, we've also opened offices in Detroit, Germany, and soon in South Korea.
Hilik Stein:
Our core technical strength though are the AI and data science, embedded software, wired and wireless networking technology, as well as of course, cyber security. SafeRide uses AI technology to enable the automotive data-driven revolution. We see a world where automotive engineers can easily use AI and specifically deep learning to leverage the vast data available in cars today to make connectivity and mobility more secure, reliable, and enable revenue. Our product create value for automotive customer by prevailing for providing early detection and prevention solution that enables financial gain.
Tyler Suiters:
So Hilik, you used an interesting term, which is vehicle health management. It sounds a lot like digital health for humans, but this is something it seems that all of us who are driving or riding in cars in the short term future need to think about and be very aware of.
Hilik Stein:
Absolutely. It all starts with the complexity of today's vehicles. The modern car is a very complex machine. There're numerous paradigm changes in the automotive world that pushed this complexity. Just to name a few, it's electrification, software defined car, convenience features connectivity, and there's just so much data and so much software embedded into the vehicles today. Modern vehicles have dozens of computers or ECU, electronic control units in them and they have hundreds of sensors and controllers. Each one of data is becoming more and more complex. Traditionally, most of the function of the vehicle were performed by simple microcontrollers, each dedicated to a very focused and well-defined task. This allowed very lean hardware and software prod, which was easy to maintain. Lately with all of the revolutions and the changes to the vehicles architectures consolidating many of these functions and installing stronger, much more complex microprocessors.
Hilik Stein:
This is a completely different game. It involves hundreds of millions of lines of code, advanced operating systems, enhanced in vehicle and out of vehicle communication and a lot of other enhancement and changes. This new, very complex centralized microprocessor based vehicle architecture provide also much more functionality, including managing over the air software updates, providing cybersecurity functions and providing a platform for new software services in the future. All of these computers are also network together and sharing data. So those so much data that we can actually play with. Modern vehicles today generate tens of gigabytes of data every hour. Future autonomous vehicles will generate more than 30 terabytes every day. That's huge amount...
Tyler Suiters:
That's worth underscoring Hilik. What you just said, I want to make sure I get those numbers right for our audience. So today our vehicles are generating tens of gigabytes every hour, in terms of data. We're heading towards 30 terabytes every day?
Hilik Stein:
Yes. Even if you look to the back and like five or 10 years ago, it was again several order of magnitude, less than today. So it's really rising exponentially. Now another major disruptive change is the software defined car. When I was living in the Silicon Valley, I had a friend who drove a model Haas. When he purchased it, it had very low level driver assistant functionality, but apparently all of the hardware and sensors and everything required for it to become autonomous was already under the hood. In time, Tesla was able to develop it's out of pilot functionality and offer it as a downloadable software update. Imagine that you just woke up one morning and discover that his car can just drive itself and all of that software update overnight.
Tyler Suiters:
Let's talk about your newest product, vInsight, brand new under the heading of vehicle health management. So what is vInsight and talk about the application as well, Hilik. The fact that this is for OEMs and the auto sector as well as aftermarket issues.
Hilik Stein:
Of course. So vInsight is safer, it's AI based vehicle health management, or VHM in short. It's a platform, we developed it for automotive OEM, still want suppliers, aftermarket telematics vendors and fleets want to move from just being able to detect vehicle health issues, to also being able to predict them and to determine their hood cost. The vInsight platforms includes both the cloud-based VHM development tool, as well as inference engines targeted for both in vehicle inferencing, as well as cloud-based remote AI based health monitoring.
Tyler Suiters:
So that's interesting that the corrective feature is an obvious conclusion. I have a problem, vInsight helps me to diagnose it, discover it, cure it. Mirroring really the health process, but the predictive element is much more innovative. This is where AI, artificial intelligence plays a paramount role in vInsight, correct?
Hilik Stein:
Absolutely. So the key is machine learning or most specifically deep learning. Deep learning technology is able to conquer the vehicle complexity and really glean insights from the tremendous amount of data that's available in the vehicle. vInsight is an AI based VHM platform that delivers three in major advancement to vehicle health management today. First it enables deployment of deep learning algorithms, onboard vehicle of where access to massive data is available, compute is more limited and you need to be in running in real time. So the AI engine first learns the normal behavior of the vehicle offline through the cloud and then component. Then it monitors the data in real time while the vehicle is on the road to identify deviation from the expected behavior. Think of it like a diagnostic tool that usually you see in the dealership, but this one is riding with you in the car.
Hilik Stein:
Second, it also brings together the expertise of the automotive engineer with the state of the art AI technology that is packaged within. It enables automotive engineers who aren't AI experts to implement VHM solutions that combine advanced AI mathematics with their own mechanical and electrical domain expertise. Third, it simplifies and automates the entire complex process of AI development using a rich library of VHM specific algorithms and models provided with the platform. It has a collection of ready-mades set of models, algorithms, and pipelines tailored and tested for numerous vehicle systems and function. All you need to do as an automotive engineer is to load the template, hook it up with your own vehicle's available data inputs, and you have your own custom VHM solution with minimal effort.
Tyler Suiters:
Let's hone in on that and point to Hilik. That is the efficiency aspect of this. For the engineering side, you mentioned auto engineers, you as CTO, perhaps not surprising to have a deep tech engineering background as well. What is the real benefit? The unique benefit of a VHM that SafeRide delivers from the engineering side, what it means to those engineers who are working on cars at the present with an eye on cars of the future.
Hilik Stein:
So the engineers that are working on these vehicles are primarily mechanical engineers, electrical engineers recently, maybe even software engineers. AI is the next frontier and it's something that requires a lot of expertise and a lot of training and a different background. What vInsight platform is aiming to do is to make all of that readily accessible for them and allow them to hit the ground running with a framework or a system that already has the solutions baked for them. So they can create all of those solutions much faster and cheaper actually.
Tyler Suiters:
No, that's an interesting point to raise Hilik. The idea of cloud-based solutions, but also onboard solutions. So what are the benefits of vehicle health management software actually in the vehicle versus a cloud storage alone?
Hilik Stein:
So certainly VHM software can be deployed either onboard the vehicle or in the cloud, or depending on data access and available computing resources. The benefits of integrated VHM in vehicle is that it can access much more data in real time and therefore provide better insights. A new vehicle architectures today have central computing and communication unit like the service oriented gateway solution from an expert semiconductors that brings together system-wide data access and computing resources, which enable deployment of advanced and real-time VHM algorithms, including deep learning on board the vehicle.
Hilik Stein:
A key part of the vInsight platform is vInsight edge, which is an embedded VHM runtime engine designed for gateway models, domain controller, and telematic models. It enables real-time inferencing of the train VHM algorithms that were created by the vInsight developer in the cloud. vInsight edge is pre-integrated already today with unexposed F32G, a vehicle network processor. Another component that vInsight edge has decay includes is CAN Optimizer. CAN Optimizer is SafeRide's data optimization compression solution. It enables efficient data collection from the vehicle for training and for remote diagnostics. It provides a 96% loss, less data compression, which is six times better than any other compression solution out there in the market.
Tyler Suiters:
So who are your primary customers? We talked about the OEM market, we talked about the aftermarket, who is the real sweet spot for say fried that you're trying to reach, Hilik?
Hilik Stein:
So our primary customers are vehicle manufacturers. They use vInsight to develop in vehicles, smarter diagnostics and prognostics software based on AI technology to establish remote monitoring diagnostic services for their customers, both private car owner, as well as commercial fleets. Other important customers are commercial fleet management companies who use vInsights to add advanced health management capabilities to existing fleets. By using SafeRide technology our customers can not only detect current vehicle health issues, but they can also predict future problems and schedule maintenance before the vehicle breakdown, which makes it much cheaper to fix. vInsight can also help the technicians at the dealership determine the root cause of the problem and make the correct repair quicker and cheaper.
Tyler Suiters:
One of the interesting aspects of your development process Hilik is your engagement with your customers, the back and forth sharing of data both raw and anecdotal as well. I don't want to forget the subjective nature of engineering world. What are you hearing? How is that driving your development, that regular feedback with the people that are your customers and who are experts in this field as well? Then second part of that question, I think we'd be remiss not to ask how the pandemic has affected that process and the efficiencies that drives.
Hilik Stein:
So in the past year we've been working very closely with several lead customers, both in North America, as well as Europe and worldwide in order to test and mature our VHM technology. The feedback so far is phenomenon. Our team expanded and we keep growing and we registered multiple patents in this area. I believe that we've established a strong leadership position with our technology and innovation, and we see great momentum and strong need for products in the market. COVID-19 certainly presented the challenge early this year and we are thankful to our partners and customers for continuing to work with us through these challenging times. Well, certainly committed to supporting them and we work with them very closely during these times.
Tyler Suiters:
So Hilik, we've talked a bit about the presence and the short-term future for vehicle technology and the move towards self-driving. Where is the future in the mid or long-term in your minds? Where is auto technology going and what do you envision as SafeRide's role in that future as it plays out?
Hilik Stein:
So the future is more, it's more autonomous driving, more electrification, more connected services, and much much more data. SafeRide uses artificial intelligence in order to harness this vehicle data, in order to solve challenging problems and bring insights to our customers to help them to improve profitability, prevent reputation loss, and keep the customers happy. As long as there's more and more data and more and more functionality, I think there will be more and more use cases for harnessing AI technology to compliment what the engineering are doing presently.
Tyler Suiters:
Well, it's an excellent point about electrification Hilik and that I think more and more of the conversation around the future of vehicle technology is about self-driving of course, but electrification is becoming a higher priority, or at least it has higher visibility as a key element moving forward. I know in Israel, this has been a priority at the government level and certainly at the level innovative level where you've been to get to an electrified vehicle fleet on the consumer side. Can I ask you for prediction as to when electrification really takes flight and maybe we hope that [inaudible 00:21:01] get to the majority of drivers globally using EVs?
Hilik Stein:
Well, I think that's a very interesting question. I think the answer is actually geographic. If you look at North California, Silicon Valley area, I think electrification is almost everywhere. I think there are more electric vehicles than actual non electric vehicles. But that's a unique geography. It also relates to how people are living. It's mostly single family homes, where you have access to two charging ports and everything. Just to give an opposite example, Israel is mostly a shared apartment buildings. With apartment buildings, getting the infrastructure, to be able to charge your electric vehicle is much more complicated. You need to go through the rest of the tenants, you need to get consent, not everyone has a dedicated parking spot. So that sort of things really delays things. So I think that the delay in adoption of electric vehicles in places like Israel is not technological, it's mostly really, I don't know, social, maybe.
Tyler Suiters:
It sounds almost practical. Where can I park? How far can I drive on a charge?
Hilik Stein:
Yeah, no you can't even really charge your car at home because you don't have a dedicated parking spot, you don't have your own meter, you can [inaudible 00:22:53]... Even if you have a dedicated parking spot, you need a consent from your neighbors in order to put in a charger. So it's much more complicated.
Tyler Suiters:
Back to the larger topic at hand Hilik. That is the most immediate future for vehicle technology. What are SafeRide's plans at CES 2021 and all digital show, a different format? How are you all leveraging that? What are you showing off to the audience?
Hilik Stein:
Yeah, so CES is certainly very different this year. I'm used to ending CES every year with soviet and a new pair of sneakers. I guess this year, it will be mostly my fingers that are going to get [inaudible 00:23:45]. We have a wonderful virtual booth here at CES, where we will be showcasing the vInsight platform. The demo that we'll be showing, will provide an example of how vInsight can be used to detect and identify problems in the car engine, even though the check engine lamp isn't illuminated on the vehicle. Now what we are demonstrating is how to create the engine monitor through the vInsight platform, and also how it's used in the vehicle on an ECU to detect an issue with the air to fuel ratio on a real car.
Tyler Suiters:
What is your elevator pitch? So to speak Hilik. It's something that I'm sure you've delivered thousands of times each time you've attended CES in Las Vegas in years past. Yes, it's different this year, but what is the one takeaway you want to give your audience? I mean, you never know if you're sharing it with a journalist with a global audience or a CEO that's considering takeovers or a VC investor. What is the one takeaway you want about SafeRide this year?
Hilik Stein:
I think that the industry has a health management, it have diagnostic services, but the industry is going through a major shift towards integrated vehicle health management inside the vehicle to achieve, to go from diagnostics to pro diagnostics, to eventually even self-healing car. I think our main role in this journey that the industry is taking is to actually help the OEMs and the Tier 1 suppliers and the rest of these players to really make this transition or tail and make it much easier and much simpler to adopt. Does actually a standard that the SAE had released last year, it's the `JA6268 standard. It's detailed six levels of VHM capabilities. Going from a level zero being, no VHM [inaudible 00:26:09] capabilities, all the way to level five that's actually self diagnostics and self-repair. The industry today is mostly at level one and two, which are diagnostics. Our goal and our role is to help the industry to go to level three and beyond.
Tyler Suiters:
A veteran of maybe the two biggest innovation hotbeds in the world, Silicon Valley and Tel Aviv, Hilik Stein is now CTO and co-founder of SafeRide technologies. Hilik a fascinating conversation, thanks for your time and best for an outstanding CES, 2021 for you and for SafeRide technologies.
Hilik Stein:
Thank you. Looking forward to seeing everyone online and hopefully next year in person.
Tyler Suiters:
That is a wrap for this edition of CES tech talk. As you know, CES transcends what is the traditional tech industry. Companies can use CES as a platform to show how they're embracing technologies and evolving their businesses, whether large or small or anything in between companies have a platform at CES. We want you to be CES ready. So subscribe to the CES tech talk podcast and that way you won't miss a single episode as you're getting ready for the big show. Speak of CES 2021 runs January 11th through the 14th. The latest announcements and exciting news are all ces.tech. That is CES.T-E-C-H. As always none of this is possible without our true stars, executive producer, Jennifer Drogas, assistant producer, Kristen Nemeroff and our senior studio engineer, John Lindsey. I'm Tyler suiters. Let's talk tech again soon.