Mercedes-Benz’s Research and Development centre in Bengaluru has evolved from a hardware-focused engineering hub into one of the company’s key global R&D centres, with teams working across software, ADAS, electric powertrains and physical vehicle development. In this interview, Jörg Burzer, Member of the Board and Chief Technology Officer, Development & Procurement at Mercedes-Benz Group AG, discusses the centre’s expanding capabilities, the role of AI in vehicle development, the future of autonomous driving, and why he believes every Mercedes-Benz has “a bit of India” in it.
A lot of global auto OEMs have large R&D centres in India. Beyond cost, what are the benefits of having such large setups in India
Yes, it’s not only cost; you see, the competence level is so high. If you look at our centre, we will be celebrating its 30th anniversary here in India, and over these years, it has only grown in size and capability. We started with hardware, but now it’s massively into software, infotainment and autonomous driving. The team in India is contributing heavily on ADAS, for instance, and we are working on Level 3 and Level 4 with the team in India. So, all in all, I would say it’s competence, speed and agility that are good here.
Is the Mercedes development centre mainly a digital unit, or is physical development also being undertaken here?
Oh certainly, we have lots of hardware development. There are many test benches, and you’ll see lots of physical testing and development taking place here. Even beyond powertrain, there’s a lot of physical development taking place along with our vendors here too. And then, on electric vehicles, we’re working on the powertrains and batteries too. And yes, on the digital side, there’s lots of AI work, not just on the product, but also on process development. So yes, the centre is a full-scope R&D centre, a blend of the digital and physical. For example, we do a lot of work on the simulation of sheet metal forming and welding here.
Besides engineering, does the India centre also do design and styling work for your products?
On the design side, at Mercedes, we have this philosophy that we are creating a global product, and so we have a global design. This might change a little bit – obviously, we are adapting our design, which might be enriched by certain market trends – but for us, it’s very important that it is a global design, so we have one design hat. So, our styling team is based in Germany, where we have our design centre, and that’s important for us to maintain the global approach.
Software is today a leveller, even phone manufacturers can make a car today. Do you view this as a threat?
I see this as an opportunity because, for instance, you can develop a software stack for ADAS, but at the end you have to apply it to hardware. One interesting bit, to my knowledge, is that we are the only OEM doing a Level 2++ ADAS with a combustion engine, otherwise you only see this on EVs. To do this in a combustion engine environment, you know, with things like a transmission, it’s more complicated, so that’s where our expertise in traditional technologies, along with our developments in software, has an advantage. And of course, we have tremendous respect for our competitors and we are seeing what they are doing, but for me, I think if you do this mix right, you will be at the forefront of development.
Past Mercedes cars stood for robust engineering. What do you think a Mercedes 10 years from now would be all about?
Obviously there will be tech, but there will be more. I can give you a good example, one of the most important characteristics and values of a Mercedes-Benz is safety. We do plenty of crash tests every year for our cars that we launch, plus there’s a lot that we are doing with autonomous driving too, for instance, what we are doing with Nvidia. And here we have a lot of redundancy. We have a traditional software stack with massive amounts of info like what is a stop sign in every different market, for example. But there’s an end-to-end stack where the software is trained by the behaviour of the driver in traffic, but what we are doing is we have both stacks, so if the end-to-end system does not know what to do, then it can fall back to the classic system. So, we create a safety environment, where we know there is always a solution.
On redundancy, what is your view of a wide sensor stack like radars, lidars and cameras, or is AI with cameras only fine?
We are using AI, yes, but we have radars too. We have to use AI to create efficiency in the simulation and training, but you have to also think about the sensor set, and for me, I am convinced you have to have a multiple sensor set, not just one type. So, like on the S Class, we have multiple cameras, radars and ultrasonic sensors too.
Does the India R&D centre develop products or platforms independently of headquarters, or is it more synergetic?
For us, it’s very important that we do everything as a team. So, for me personally, it’s very important to have a global perspective in R&D because the innovations and technology solutions are so complex that you can only master them as a team. There are a lot of competencies, but there is not one single team in the world that can say, ‘I am the one that can create a platform.’ It’s always teamwork, and if you look at all industries, automotive engineering is the ultimate for teamwork. So for us, R&D is one big team. And with regards to the India centre, I love to say that in each and every Mercedes, there is a bit of India in it.
