In this Top Floor interview, we hear from Sergey Malygin of Soda.Auto. Sergey is at the forefront of AI in the automotive industry.His software significantly reduces time on production and enables various types of vehicles to be made with a few clicks.He also outlines how he used learnings from one industry to transition to another. For more information on their services goto https://soda.auto/
Connect with Sergey Malygin on Linkedin: https://www.linkedin.com/in/smalygin/
We hope you enjoy this episode! Give it a like and subscribe if you'd like more content like this :)
From
The Top Floor Team
#ceointerview #businessleadership #businessleaders #ceo #ceotalks #businesstalks #ceos #ceosdesk #ceoadvice #podcast #podcasts #podcastshow #podcasting #podcastclips #podcastseries #thetopfloor #topfloorpodcast #foryou #foryoupage #fyp #fypシ #fypシ゚viral
So hi everyone, I'm Jamie Grell, the host of the top floor CEO podcast. We've got a very special guest today. We've got Sergei Maligan from soda.auto. Now Sergei's had an amazing story he's going to share with you today and there's lots you can learn from him. So firstly, he's successfully transitioned from one industry to the other and also what he's doing now at soda.auto, he's transforming the automotive industry using AI and software engineering. So we're going to start firstly, so Sergei, so welcome to the podcast, delighted to have you with us. Hi Jamie, my pleasure, great to be here. So just firstly, just start, I want you to go back to maybe the early days that's relevant for people listening to look at your story. Just tell me firstly, why did you get into this particular area of expertise? Yeah, well it's an interesting story to tell and I started a long time ago and when I described my experience I say that it's two big parts. One big part is telecom and then another big part is automotive. In telecom I started early and I launched mobile voice operator. I started and deployed a new product for smart traffic management within the network operator and it was all really great. But then after 10 years I've got an interesting proposal, hey can you help me to make an electric vehicle and can you help me with operating system for that vehicle? That's how it sounded in the very beginning. I said well it sounds really interesting and I never did that before but never done this before and yeah let's do it, why not? That's how I ended up in automotive and that's the domain I want to say. I love it here and that's where I worked for the company called Arrival for many many years and nine years in total started with a small team and very soon from that I realized that it's not just about one operating system, it's all about the entire ecosystem of different components and system working together as a whole vehicle and in my ending part of being with Arrival it was 750 people in total. We were looking after all the different systems we created the ecosystem of different building blocks, hardware and software for which Arrival could make different types of vehicles, vans, buses, passenger cars so it was a huge nice journey so far. Then in parallel to that there was another project called Robo-Race where we launched a vehicle which could run autonomously at a racing track at a high speed so fully driverless 11 months it took us just 11 months to make it. It was a racing car running in Marrakesh with informal e-environment concrete barriers and all of that with the speed of 200 TPH for like 12 flying laps fully autonomous nobody inside so in a nutshell that's my story. What is unique about it is that I saw so many things in telecom prior to joining automotive when I switched to automotive I started to see that I actually took the time machine to go back in time and now and now the same things are happening in a different industry. So what things that you learned in telecoms could you transfer to automotive? Look in automotive so many things are happening now around software defined vehicles and in 2014 when I just started in automotive I started to actually create an electric vehicle based on software defined vehicle approach from the very beginning because I simply didn't think that it can be different but in automotive at that point of time it was absolutely nowhere close it was so different so when I was in telecom I witnessed this transformation we all remember Nokia and we all remember what happened in 2007 when Apple iPhone 3G was introduced then 2008 Android was introduced by Google 2011 whereas Nokia lost all their positions so it took just four years for software defined and software oriented players to take the leadership position on the market and the same thing is happening in automotive okay it may take a little bit longer not four years it can be five ten years but this is happening and we're right in the beginning of this journey and some of the companies are ahead of time like Tesla and some Chinese examples some of the companies are still trying to catch up but it's happening this is the moment when it's happening and what is great for me is that I know exactly what is going to happen and with soda.auto being focused on software defined vehicles we know exactly how to help many many companies within this transformation so do you supply to people that make cars is that what do you supply your services to so yes our product is the unique method which is not a theoretical SDV not a theoretical software defined vehicle approach but a very practical framework so what we do we provide we provide the ready to use library of software as well as single tool which can substitute 15 different tools which were previously used with a unique method which can help any OEM organization any automaker to create the vehicle with software defined approach in mind at least two times quicker and four times less cost that's our promise that's what we confirmed throughout our current customer base and that's what we do how do you do that so okay very good question so look it's it's actually quite simple if you want to make things faster you need to change the mode you're doing that change the way you're doing that so one tool a lot of automation which reduce amount of integrations you need to have reduce amount of different systems you need to just put together and apply automation process inside then you reuse the when you reuse the existing building blocks you don't have to develop them from scratch it's it's the second pillar of that and third is that you apply AI because when I founded the company last year it became so clear to me I had a big meeting with my team and I said look guys actually starting a company 2023 and doing anything without AI is a crime I mean it just no sense we we need to apply AI in every aspect of what we do it includes our tools it includes our processes so the way how we operate internally as the company as the business and of course it includes the software and the way how we deploy the software to our customers so a big pillar of what a so of what solar is today is AI artificial intelligence and we deploy it in every part of our tool chain so I'll give you one example in a traditional process what you do you you start with your requirements you define them manually looking into the regulations following functional safety cybersecurity regulations as well so you spend a lot of time just making your requirements then what you do you make the implementation then you make the testing happening then what you do you write your tests you are basically taking the requirements and you make another lines of just a text which represent a test you're going to perform and then what you do you convert this test into the executable test something which you can really run and check if it's working or not so what we do we use AI to help us and all of that so with requirements we take the regulations and we make the requirements in automated mode so it's just 15 minutes to review then we make tasks automatically from the requirements with the same kind of pattern you just spend 15 minutes to review rather than a few days to to make your testing problem ready and then you automate the execution all of that can be done in just minutes compared to days and months wow so the IO is just cut the cut the time dramatically that's kind of what it's done absolutely so the way we position ourselves and the philosophy and the vision behind everything in soda is three main pillars SDV software defined vehicles yes it's a must then digital twin another concept which we follow everywhere and AI by fusion all of this for fusion of all of this free technologies together we achieve ultimate levels of automation and efficiency so the so you talk about the AI makes it a lot takes less time did but it also do things or you know that stage yet what humans can't on involving the technology again so you mentioned that what the AI does is speed up yeah the solution so you can do what humans do but quick up as can it do yet what humans can't do yeah are we not there yet well I would say it's a good question AI can help in some things with human either can't do or don't want to be doing or it's not even a human limitation but the technical limitation one example is negative tests when you work with vehicles the last thing you want to be doing is to make like too many crash tests or to make too many negative tests where everything is not working and you just you know spend money and waste your money on more and more and more prototypes so what we do inside of our system is we inject the faults and we make this negative testing with the help of the AI a lot which can actually try to think of all negative cases for us and we test all of that in a virtual environment in a digital environment first so we don't have to repeat this test in full scope in real life so it's not exactly what you mentioned but that's one of the examples where AI is so beneficial in terms of human limitations well I can say that when AI gives us some output sometimes we're quite surprised with few bits which well I didn't think about it so it's a different perspective so it's just you know extra input of information still as we said human review is important but AI can suggest the things which you would not normally consider and when it comes to safety and security that is a real value to be added into the process how safe is driverless cars driverless cars well it's an interesting question I would say the technology is there I would say that if we look at all bits of technology which is available it it can be safe enough the the only bit which is missing a little bit is the final integration and also the ability to test and validate it in all real use cases across the globe because when you when you take a human I can drive in the UK I can go to Asia and try to drive there which is becoming a little bit more challenging but still doable and the same with AI you need to train it with so many different conditions with a level of technology we see today it becomes possible because you even can have a synthetic environment to train it before you go to to real life again and integration it together it's it's it's already there we see many companies who are quite good with that and I like the example of wave in the UK they're really doing great and they're this next generation of company we see the latest levels of innovation achieved by OpenAI and the the smartness which AI can can can get to my favorite example here is when you ask AI to do something and you ask it in December and you ask it in November and then surprisingly in December you get a different result why because AI learned that actually in December you can be slower because you're preparing to Christmas so it learned not just you know the informational bits from from its data set but it learned human behavior so when you see how AI can learn from such elements you say wow and that's that's how smart actually AI is today so all of that coming together is is is here so it's it's it's it's with us the biggest thing for AI in terms of the safety is more testing on real roads and ability to validate it in real life environment and if we take China for example they accept the risks so they do testing just on public streets and roads which is not the approach I'm I'm proposing here but I'm saying that when you can allow this this of course helps to take the technology quicker to the level where it can work in real life so the missing bit is not the technology today the missing bit is a social effects is the regulations and legislation and just the public perception of the technology as well okay so just talk about the type of businesses that you help just describe what they do in areas that you can help in to look as I was saying in soda we have a practical solution for anybody who wants to make a vehicle who wants to make an autonomous driven vehicle any vehicle just calls is there any type of vehicles that boy any vehicle it can be a van it can be a bus so actually we are not limited to automotive purely as well so it can be robotic environments so it can be vehicle in a wider as coal so it can be agriculture it can be marine so our tool chain our approach and the library of reusable assets actually is equally great for all of this application so anybody who's making a vehicle in a wider way of thinking about drone as well our space is covered and wants to get it faster to the point where they have something really working take the tools take the hardware take the software and integrate it very quickly is and at a very good price level as well we offer very competitive prices that's our very natural customers okay so basically you look at some of us that is developing a vehicle they want to make the process quicker and also is there particular features you can add like you said you said about driving this you can add that is anything else that you can add that will make it you know unique and something that can't do themselves well yeah it's a good one so let me describe this a little bit so it covers all of them it covers very basic functions so every car every vehicle needs to be able to just to drive to steer to be to charge then you need to have like seat belts lights wipers all of this you know very basic things we have all of them on top of that we have white advanced things such as torque vectoring and advanced traction control so you can do the things like tank turn this is actually one of the interesting examples we achieved with one of the customers who made the car spinning with four motors configuration each motor per wheel and what is quite remarkable with is that with our approach we achieved that in just seven days so we made a very complex thing which very big brands practice for many months before they could demonstrate it in public we made it in just seven days that's one of the examples another example is aye climate control think about your vehicle experience you get into the car it's a hot day around first thing you do you turn on to your AC right your your climate control but then you need to calibrate it you need to make sure it's good for you and then and then finally it's still not blowing right for you imagine a different use case you have your smartwatch and the car actually knows if you're hot if you're running or is just hot and warm outside and then before you get to the car it is prepared for you so it knows your temperature it knows the outside temperature and it knows the temperature inside of the car so it can prepare your car to give you a perfect mode for your climate control to be exactly right to these parameters that's what we deliver as one of our features and this is what is pretty unique and we can offer it with pleasure to to our customers excellent okay so can you talk about some projects that you're working on so look with with the projects we are working on right now we are looking forward with the progress and innovation in the entire automotive domain so right now we already automated the software documentation process so what we did we applied aye to create documents for us so we don't spend time on that we just review and at the same time we are planning to extend it to use aye to actually make software for us so it sounds scary but that's why we keep a human in the loop as well so we review everything aye suggests to us but imagine the process where your new software again is done with the help of aye rather than you writing it for many many months et cetera et cetera even today we experiment with aye system like approach with multiple aye agents with mabs approach aye systems work together and think of a meeting of different people in the room what normally they do they kind of discuss then they agree negotiate they disagree and then finally they came up with something in common in the same way when you design a new product you can one person makes requirements your system engineer another person is responsible for cybersecurity the third person is responsible for safety then they meet and then they make your requirements set like non conflicting and balanced so what we did we literally made the same thing with aye agents so one aye agent represent your safety expert another aye agent represents your from your cybersecurity expert and then the third one is your system engineer and then they talk and then they find a common ground and then they provide you with the artifact of what they kind of were able to make together and you know what it's amazingly good that's fantastic sounds phenomenal so just tell me about your vision surrogate so firstly for soda the auto in the next say five years and also say the industry is a whole just just share what what people can expect and how how you're seeing it i will possibly start with a second so industry as a whole is making a shift from hardware defined to software defined aye is jumping in as a huge assistance and actually a huge leading force in the future automating most of the operations and that leads to more efficiency better products and better safety actually and security for for all of us so that's the transformation the industry is going through with soda what we do is we're helping this transformation to happen we're pioneering some of these technologies and our ultimate vision is to make vehicles possible to be made with just few clicks that's a mission that's what we are working on totally been the factory fantastic that that'd be phenomenal so there's anything else so you can share to people that are interested in the technology interested in the future that that would help them learn a bit more about this exciting space well i would say that the key thing is that so many new things are possible today with the with the revolution of aye with digital twins and actually with their fusion because with this smarter agents which learn the object and can suggest the changes of an object you eventually come to a point where aye can can generate so much so with this technology is in place everything which wasn't possible before becomes possible so let's be creative let's let's try to imagine and unleash our imagination and let's apply the latest technology to achieve all of that everything is possible today fantastic and how can people get hold of you so aye good second how can people get hold of you if there's like a someone that's looking to get this automotive technology to help them how can they get hold of you to find out more absolutely so there is our website soda dot auto and also my personal linkedin page third game a league and you can quickly find me and i'm super open to all the conversations around the space and everybody who wants to learn something about what we do be a partner to us or be a customer to us you are more than welcome to join me and my team we are super passionate and ready to help to everybody i bet you that's soda dot so that's s-o-d-a dot-a-u-t-o exactly excellent okay so okay it's been an inspiring conversation it's a very exciting technology it's all very new it's got a eye at the centre of it and it's we're really looking forward to seeing where this space goes and thank you for sharing your insights today because i'm sure there's a lot of people not watching this it's going to really enjoy it and and let them know what the possibilities are thank you very much okay thank you very much for having me jimmy have a great day and thank you everybody thank you
