AVL's Reimagine Mobility Podcast

Middlelayer Mobility Software Development w/ Vivek Jaikamal

• AVL, North America

Share episode

This week, we're venturing into the future of transportation with a special guest, Vivek, a leading innovator from Apex.AI. Together, we'll explore how cutting-edge technologies and visionary ideas are setting the stage for a revolution in mobility. Vivek's unique insights and experiences at Apex.AI provide a fascinating glimpse into the advancements shaping our travel tomorrow. 

 

Apex.AI aims to provide a robust, safety-certified middleware framework for companies to build their applications on top, allowing them to differentiate their vehicles based on the Apex middle layer. This allows OEMs to differentiate their vehicles based on the applications, regardless of the underlying hardware and operating system being used. By providing a safety-certified middleware framework, Apex.AI aims to provide a reliable platform for companies to build their applications and differentiate their vehicles.

 
Your journey into the next era of transportation begins here. Don't miss this insightful conversation that bridges dreams and reality in the mobility sector. Tune in to the Reimagined Mobility Podcast on your favorite streaming platform and witness the future unfold with Vivek and Apex.AI Join our community of innovators, thinkers, and explorers by following us on social media and sharing your thoughts on this episode. Ready for more? Subscribe to our series for a regular dose of inspiration and stay at the forefront of mobility innovation.

 
 Vivek is the Director of Business Development of Apex.AI. In his role, he leads the business development activity for Apex.AI in North America, enabling the transition of automotive and non-automotive OEMs to software-centric development by adopting a truly cutting-edge framework and methodology derived from the ROS open-source community.

Vivek is a pragmatic engineering and business professional, passionate about leveraging the power of software to solve real-world problems. During his early years in the automotive industry, he worked in the General Motors team that developed the first stability control system to make cars safer in tricky road conditions. In his subsequent roles at Bosch and AVL, he helped several automotive OEMs reach their goals of developing safe and reliable mobility solutions using innovative approaches, advanced development tools and safety-certified operating systems. During this time he also co-authored several SAE papers highlighting those methods and tools.

Vivek obtained a Management Certificate from the University of Michigan Ross School of Business, a master’s degree in engineering from The Ohio State University and a bachelor’s degree from the Indian Institute of Technology. 

https://www.apex.ai/

If you would like to be a guest on the show contact: namarketing@avl.com

Welcome to the Reimagined Mobility Podcast series here from AVL I'm here, my old colleague Vivek. Back in our days. Thanks for joining me, Vivek. You just got out of this. Before we really jump into how we reimagine mobility with what you guys at Apex are doing where you're working at today. Give me a little bit of a background so that our listeners understand where you're coming from, what your background is, what the company you're working for right now is doing. And then after that, let's jump into it. Great. Thank you so much. Yeah, it's my honor to be talking to you guys today regarding give me regarding mobility. I'll give you a brief background of myself. So I'm an engineer by training. I started my career building simulations and for testing of brake systems and steering systems identify. In the early nineties, I got really intrigued with the possibility of making cars handle extreme situations, like in a slippery roads and high speed turns safely and elegantly with the help of embedded software. So that really sort of you know, put myself on a trajectory to designing and writing and better software for the early airbags and traction control systems for the GM vehicles in the nineties. And that was a really exciting time. This was, you know, we were doing this quite early. We had very primitive tools and methods back then. And then I moved to Etailz, which is, a division of Bosch. And at the time the software complexity was already going through the roof in terms of the not available microcontroller capability back then. And Etailz, was sort of a pioneer in industrializing the art of automating the creation of high quality and better software through tools and methods that were developed at Bosch. So I had the opportunity to bring these tools and methods to all the big OEMs in the US, Europe and South America as the industry with adopting, you know, more and better software and and better technology in these vehicles. And what attracted me the most was the opportunity to work with some of the brightest people in the industry over two decades and watch the automotive industry evolve from being purely mechanical, as you guys know very well, to becoming more mechatronic with the role of the software was increasingly becoming more and more important, and that was already as hot as only accelerating the last 15 years or so as we all know. So after it, as I joined AVL, I had the unique opportunity to launch some of the most advanced simulation and testing tools for the next generation technologies that were coming on, like driver assistance systems and driverless cars. And during that time, electrification of the automotive industry was also gaining momentum. And I saw firsthand the new electric vehicle manufacturers how they were building vehicles, which was quite different from the way the established OEMs were were doing it. I then had an opportunity to work for a small startup which was building driverless, technology, both hardware and software for public transit vehicles like a small 6 to 8 passenger, low speed shuttles and up to 25 passenger vans. So this this is where I saw firsthand how many cities, campuses, communities were experimenting with the idea of adopting driverless technology to address some of their needs in cities, especially around safety, and especially during COVID, when it was becoming more difficult to have people move around in a congested environment. So during this time, I also realized that for these types of vehicles to become successful, you need a very robust regulatory environment as well as acceptance by the public. The safety was becoming a top concern for everyone. And at this time, in the middle of the last year, 2022, I got the opportunity to to lead the Apex A.I. business development activities in the US. So that's where I am today. And, I realized that the automotive industry was changing rapidly again from being hardware centric to becoming more software centric. And for anybody who has seen the evolution of electronics in the cars, as probably seen, for example, you know, the dashboard displays are becoming more flexible and user configurable, the displays screens are becoming bringing in new navigation and entertainment options, for example. But it's also frustrating that once you buy a vehicle, it is practically already old because it's fixed in its capabilities for the time that you own it, which could be 10 to 12 years. So people are demanding that, these vehicles can be updated with the latest safety, convenience and personalization features throughout the lifecycle of the vehicle, just like your smartphone. And companies are responding to that trend. but that requires that the underlying software and hardware architecture of these vehicles be drastically changed. So this is very Apex.AI comes in, which is where I work today. So we provide the underlying software development kit, commonly known as the middleware that essentially abstracts the underlying hardware, which is the compute. from companies like Intel, Qualcomm or NVIDIA and the Real-Time Operating System just like, you know, QNX or Linux, and allows you to build a modern applications on top. So we have a couple of products like Apex.Grace and Apex.Ida that are built on open standards such as ROS, which is short for the robot operating system, which has been developed specifically for robotic applications by a large community of developers. the Eclipse iceoryx, which is a communication mechanism used to transport data on a single chip and Eclipse Cyclone DDS, which is a communication mechanism used to transport data across a network such as Ethernet. So we are basing our products based on these standards, but we are sort of industrializing them so they can be used for the safety relevant applications like automotive mobility. And we are the only company that brings the power of open standards to the automotive industry with a safe, certified and developer friendly software solution for mobility. So that's that's where we are today. Okay, good. So when I listen to all this, the two themes here are safety and software, right? So if your history and your passion, as you mentioned earlier, when you were at General Motors and then at Bosch ,AVL, it's heavily software related and it's heavily safety related. So in this discussion here, when we re-imagine mobility, what's the next step? I mean, everybody today talks about safety. At the same time, we're putting more and more autonomous features on the road that depending on which vehicles you drive, you know, I travel a lot and use a lot of different vehicles when I get to the place, rental vehicles that are trying to always have a different one. And I was always used to ADAS feature and sometimes I have to say, Wow, great job and not a case is mine is really scares me myself a little bit. Yeah. So where do you see the industry go when it's clear Software is becoming more and more dominant? You can never forget physical, mechanical and electronics, obviously, because otherwise software is not going to do anything. It's just going to be thin air. So to speak. But words to software, major and majority speaking also the safety going from here. What do you see in the next five years moving? Yeah, so I see three major trends, right? One is because we are all seeing a lot of electrification happening in the automotive industry and across all mobility applications, not just automotive, you know, even off vehicles and, and luggage handling vehicles getting electrified, which is a push to make things cleaner, more sustainable long term. I, I'm a true believer in that. I myself have a electric vehicle and also solar panels. So make sure we are charging them more sustainably. Right. and I think that's going to be the trend that gives an opportunity for vehicle manufacturers to sort of design their hardware and software from the ground up, which means they can build more software defined architectures than the legacy vehicles. So you can see a lot of the new electric vehicle OEMs actually producing products at a much faster clip in terms of, you know, revisions, updates because they have modern architectures and that is going to be the trend. I think also with legacy or traditional OEMs, they are going to re-architect their next generation vehicles. And in those terms, what I see is in the next 5 to 7 years, there's going to be a transition. for now, I think most vehicles that you buy today, have very little software updatable content. There is a some, some in there, there are some companies who are able to do updates in real time. Most of them are not. At the moment, I think, but 27, 28 timeframe, you will see more, more of those vehicles being updated, remotely features being fixed and, and new features are being provided. And by 2030 approximately, I think you will see that becoming more mainstream. So in the next few years they'll be a transition. There'll be more adoption of, you know, single different architectures, underlying architectures in hardware and software, and it'll be driven by electrification, but also be driven by, safety and automation. And the reason I say that is in a safety and automation are areas that are that can be very heavily influenced by software, safety in the engine of the, human beings being responsible for currently for driving vehicles have a certain response time and software can definitely with, you know, with high fidelity sensors you can have software respond ten times faster than humans. And that's especially critical in urban areas like in, in cities like in I'm in a small city, like in Ottawa, but even here when you go downtown, you I feel sometimes that you know, the some of the large vehicles that we see playing on the road in urban traffic can benefit from much better safety features, especially for vulnerable road users like pedestrians and cyclists. So to make roads safer, you need a better sensing capability, better reaction times from those vehicles. and that's possible through software. So from my perspective, urban safety is going to become more and more important. Yes, highway safety is also important. but I think a lot of lives lost today are in the urban environment, and we can reduce that by not only software in the vehicles, but also there are other infrastructural issues we have which we will not talk about today, but software definitely as a, as a key role to play there in the vehicles. So you can improve safety, you can improve sustainability with electrification and then automation. I think with automation, what we will see in the next 5 to 7 years is increased adoption of automation in areas where there is, repetitive tasks, there is drudgery of, you know, due to repetitive tasks like in in industrial environments, warehouses, luggage handling, things like that. Those will be automated first. Then we will see as, as automation becomes more accepted on public roads, I think we will see a bit more adoption of automation on public roads as well, but proper, no guardrails around it in terms of what can be done and what cannot be done. So automation in my mind has a much longer arc at the moment in terms of adoption. Electrification will come first. you know, urban safety and then finally automation. That's how I see it. Okay, that's good. Well, let's dive a little bit into the software, right? I mean, fundamentally, I always look at software. You have the low level software, what you mentioned before, that interfaces directly with with the chip. Right. And you have to middlelayer that you guys are working on. And then you have what I would say, what the customer can touch or at least feel and or see to some degree, which is the application layer. Right? And that's also when you talk to a software guy a whole lot more. But yeah, he's three layers. When we look at this, right? We have a lot of companies that talk about software and a really mean application layer. We have more and more companies that are talking about the importance of the middle layer. I would say there's few that really talk about the base layer software, as I call it, right? But this is certainly the critical piece to make things work. But when you look at it, when we when we when we look at the future and you mention it, the software defined vehicle, right? Dependent if it's for an electric vehicle, if it's for a for a vehicle, specifically a two 8 hours or a level four, maybe level five in the future. Software. What is the most important layer of the software and why for from your perspective? Yeah. So I think the best analogy here is our mobile device, right? So you can see on a mobile device, you get an underlying piece of hardware that you purchase. But obviously in there you have an operating system that you know, you don't see, but it's there, it makes things happen. It, you know, is sort of the, the network behind everything else. And then you have the apps on top, right? Apps are the ones that you download based on your needs and you personalize the device to your needs based on that. So something similar is happening in the automotive industry. the only challenge today is there is a variety of operating systems in place and there is a variety of hardware, technologies, hardware architectures in place, which is all good, you know, for up and on top of that, we have additional requirements for safety. So your mobile device can be rebooted with if it, if it freezes, but a car has to have a safe operating system, has to be robust and you can't afford to reboot it every every day or every second, second hour because of a problem. So there's strict safety requirements for for mobility applications. So, what we are trying to do at Apex.AI is making that middle layer, which is in between the top level applications that, make your life convenient on the road and the underlying hardware and the base levels are for the base level software is provided by companies like BlackBerry, QNX, Greenhills and others. Those are the Real-Time Operating System. and on top of that is a communication layer that essentially, allows these large amounts of data coming from different sensors to be to be transported across the, the, the, the chip chip architecture to be to be sent to the, to the sensors or from the sensors to their traders. And then on top of that is this framework that you provide Apex.Grace. Which essentially abstracts all the lower levels of the application layer. software can be written, without much concern about what's underneath. And that's what we see in the, in the Android ecosystem, for example. So you have multiple hardware providers for the Android phones, but a common operating system and framework sitting on top. And then you have personalization by the OEMs on top in terms of the look and feel of, of the Android device. So Apex.AI’s goal is to sort of become the standard framework for automotive applications going forward following that particular model, and that will make it much easier for app developers. for your convenience features like yours, you'll seat, your rearview mirror. All the other applications that you will have in the vehicle will be coming across, across different platforms, but they'll be personalized according to the OEM preference. So OEMs can differentiate their vehicles based on the applications, regardless of the underlying hardware and operating system being used based on the Apex middle layer. that's the idea. So we are trying to make sure that we have that robust, so the safety certified middleware framework available, which companies can rely on and they can build their applications on top to differentiate. Okay. So as we, as we again here together re-imagine mobility, what are the two or three things that you and your company see in the market that you would say needs to change? Right. And I'm talking maybe the supply base is not focused enough on the on the middleware that you just explained and the ability that allows you to put software onto any hardware device independent of what the what the OS is, so to speak. Again, similar to the app, right? An app I can download on almost any platform because the software infrastructure is there, or maybe again, what do you guys see that you're saying? I wish I could change maybe the mindset of the OEMs to be more focused on A versus B? And as it relates to, again, everything is related to software and safety. Let's keep that theme here from from your history and in the automotive space. But what do you see are the two or three things that if you could change today, you would change, right? A perspective of direction, focus, share a little bit. Sure. So what we see, I think this is a standard challenge in the automotive industry that we see today, is essentially adopting a modern vehicle architecture and for the, for the layperson, the meaning of architecture is essentially how information within the vehicle flows, right? It comes from, sensors like you have cameras and other, sensors in the vehicle and it goes to a certain central brain in the, in the vehicle. Currently there is no central brain in most vehicles, and so the brain is sort of distributed and that has an impact on how data flows, how, how much latency is there, how information is secured in the vehicle and as impact and all of those. And then there's a patchwork of solutions to address each of those requirements. So that particular challenge is being addressed by RE but taking a clean sheet approach, right? Having one central compute that everything comes in there so that you can harden it for security, you can make it robust, you can have you can upgrade it in one go. For example. and then you have the software development challenge. Most of the current challenges around software are related to, you know, the way software is defined in the OEM and then developed at a tier one and then brought back into the OEM for integration and then you find issues and bugs and then you have to send it back to the Tier one for, for, for fixing and then you have a delay. So that introduces a lot of delay in the overall, release of the vehicle. And then in service you find bugs and then you have to go back and fix them and then you have to get the vehicle back into the, into the shop for a software update and things like that. So there are all these business and technology related challenges. The way we propose these can be solved is working Tier Ones and OEMs working more collaboratively on a single repository, for example, the way modern software is built outside of automotive today can be brought into the automotive industry. They're working collaboratively. They are reducing their development lifecycle by taking the same approach in the prototype build as they were doing the production build. So today prototypes are built on a specific platform using specific technologies. Then we go to production. You have safety certified software, which is different from the prototype world. Now you go to re re rebuild and entire application again tested again. So we reduce that dramatically by using the ROS open standard and we have sort of hardened it for production and for safety. purposes for automotive. So you can do the prototyping, you can do the development all in one cycle and, and you can see the newer, OEMs are actually producing vehicles at a much faster clip, two or three year development timescales than the traditional OEMs because they have these processes in place and we are sort of bringing these new technologies also to other OEMs who want to adopt them. So making the development more seamless, making in production updates more seamless if you have the right architecture in place and so keeping the entire vehicle lifecycle in mind, making optimizing all these different areas of, of the development cycle. So those are the challenges the industry faces and that's how we're addressing them. Obviously there's a lot of collaboration here that is key. We are not doing everything. We are only providing the middlelayer, but we are advising our customers on, on also the right approach to the architecture and the right approach to building modern C++ based software. how to, how to do this in the cloud, how to, how to integrate everything together so that you release the development cycle. That's what we're focused on. Okay, very good. Two more questions. The first, you brought it up, right? The startups are sort of having challenged and maybe shown the traditional guys OEMs on how you can get software done faster and can launch a vehicle faster because software oftentimes now is the long lead item because it's taking over a much bigger portion of the vehicle function control or base ride. At the same time, I'm hearing, you know, people saying, Yeah, that's great, but they don't understand automotive as well. There's a lot of again, here's the word again, safety implications. If you don't do the software right, we see some OEMs talking about bringing all software in-house. Certainly some U.S. OEMs have started doing it. And we see in Europe there is one that has sort of found that their own group, which they're struggling with right now, what do you see? How do you answer the safety concern and what do you see where is this going? Is it all going? Everybody's going to bring it all in-house or everybody still using the traditional format of outsourcing and more. This collaboration that you just mentioned. But tell me a little bit what what do you see from your perspective? Where does this going? Yeah, so I think this has been an issue that, we have sort of observed over the last 15, 20 years. It's also a little bit of a cultural difference. And Europe, most OEMs have been relying on tier ones to build their software systems and to do integration in-house in North America. Have seen OEMs build everything in-house, just buying the hardware and doing the entire software in-house. And that seems to be continuing in North America. And there are advantages to that also. and then obviously OEMs, as you said, building these small groups, more agile technologies, modern software practices in small groups because you know, they have their own cultural, you know, legacy to deal with in the traditional environment. So that cannot be changed overnight. But they, they want to absorb all the new methods in this, in the newer groups. So what I, what I imagine will happen is that these technologies, the new software development processes, will become mainstream within the Williams as, as we as you talk about, making the architectures more modern in the next five years, it has to come together. You can't have two different groups doing software. And when it comes together, I think the entire, the entire company will use a common software methodology. It has to happen. You know, if you have two different methodologies, you will always have friction, you'll have all of the problems, so they will all come together. whether they all do the entire software in-house or outsource some of it remains to be seen. I think there'll be more collaboration as a, as I mentioned earlier, which means the core software will, will not be developed by the OEMs because there is no value for the OEM to build. For example, that Android OS, there is no relevance for SoCs for, for a mobile OEM to build the Android OS, it comes from Google, it's standard, they maintain it. You take it and you build the entire ecosystem around it. That gives that, that makes it looks like your phone, same thing. But for all the audience, they will take a standard operating system on the standard hardware and build, the experience on top that that makes it, that particular OEMs vehicle. So that is the, I think the overall trend. I think both, OEMs across the world will eventually follow that trend. They'll probably be two or three, market leading, middlelayer platforms like Apex in the market, and they will, they will choose which ones to use, They will all be equally performant in terms of, you know, data transport, communication, safety, security and all those things so people don't have to worry about, the integrity of the middlelayer. The user experience will determine which body, which ones are more prevalent. So the in the end, the software developers experience, the user experience is what matters the most. And that's where we think, we will end up in a few years. Okay, good. Final question for you. Yeah. What's going to be the next vehicle you buy and why so? Good question. I mean, I currently have a Chevy Bolt and I love it and I will probably continue with the electric vehicle. I'm looking at the Bolt EUV right now, which I like a lot because the Bolt EV is not available anymore, but I'm a small car guy. I live in a small cities. I would prefer a small electric vehicle which I can shuttle around 50 to 1 mile range in the city. I also have a standard icy vehicle which I use for long distance driving. So that's, you know, the whole thing about EV versus hybrid, I have a hybrid household is what I see and I see. So I am a hybrid household. I use the AC for long distance and the RV for my daily driving. So I think, the next one will be an EV. Which one? I don't know yet. My M I'll be up to replace my TV in the next six months, so I will make a decision then. Thank you. All right. Very good. Thank you so much. Later to help us here reimagine mobility from the perspective of what you have seen and what you're working in. And it's been great. Thank you. Thanks for listening to Reimagine Mobility Podcast. If you like this episode, please subscribe and tell a friend.