#ERF2025 Keynote Joerg Burzer (Mercedes Benz)

All right, Thomas, thank you very much for the introduction. Obviously very happy to be here. As far as I understood, around 1,400 participants. That's quite a lot from different countries, a very international conference. And again, very pleased to be here and talk about the transformation in the automotive industry. You all know that we are in the biggest transformation in automotive industry probably since 100 years on the product side, because we are trying to master the transformation into electromobility. But there is also transformation going on on the process side. Production processes are very much focused right now on transformation of sustainability, digitalization, and robotics, and I would like to talk about that today. Before I start, two remarks. First, I have a little bit of a cold, so if I'm coughing or sneezing, bear with me. I don't want to be impolite, just probably need a 30-second break. Second, as Thomas said, I'm an old school hardware engineer, material science, mechanical engineering. So don't expect a scientific presentation on robotics. I just want to give you insight on what we are focusing on at Mercedes-Benz. But first, I want to sell some cars. We just had the world premiere in Rome of a brand new addition to our product platform portfolio and production network. That's a new CLA. This is a symbol for the transformation on the product side. This car is our electric one-liter car for the future with a range of more than 700 kilometers. You can charge 320 kilometers in 10 minutes. Most importantly, it has a completely new MBOS Mercedes-Benz operating system, which is also influencing our production processes. We are very much looking forward to that. I don't have any contracts with me, but if you're interested, do not hesitate to contact me. We build that with more than 1,000 robots, and the launch is quite soon in our plant not far from here in Rastatt. So it's a German-built, made-in-Germany car launching in the second half of this year. Talking about production, this is our Mercedes-Benz operations network. We have around 80,000 team members at MO, producing 2.4 million vehicles a year including passenger cars and vans. We have more than 30 production sites worldwide on five continents. Our team consists of more than 100 nationalities, and we are still working together. Sometimes if you read newspapers, that's not so common, but here it's very nice. East, west, south, and Europe, we work together as one team to master the challenges of transformation. We have more than 30,000 robots in our global production network. These devices are helping us to produce the world's most desirable cars. They cover 50 production processes: body shop, assembly, paint shop. Not sure if we have robots in the legal department, but for sure in the paint process. Next week we will have the groundbreaking for our brand new most modern paint shop in the automotive industry, also not far from here in Sindelfingen. We're investing more than 900 million with our partner Dürr, also a sponsor of this conference. We are installing 100 robots in this paint shop. We have them in orange, green, yellow, blue. They are all connected to our digital production ecosystem, MO360. We are moving up to two terabytes of data per day to the cloud. Robotics and data are really connected. The load applied is up to 1,000 kilograms, and tolerances are around 500 micrometers. That's the fleet of robots we have. We have a history of using robots. Another OEM introduced the first robot in automotive production in 1965 or 1966. In 1971, we installed the first welding transfer line in our body shop. That was the start of using robots in production. But now, if you look in our body shops, you hardly see any team member, except some maintenance. It's basically a robot environment. Coming back to humanoid robots, how they are impacting production. Data and robots are very connected. Progress from 1970 to 2000 was the introduction of machine learning in robot applications, introducing sensors so robots could see, collect data, and use data to steer and control. We established in 2020 a digital production ecosystem called MO360, Mercedes Operations 360 degree, where we collect all data in our cloud system Microsoft Azure and apply algorithms. We also use the Nvidia Mercedes Omniverse to plan our factories. In 2022, we decided to allocate all our competences in one location, Berlin, our oldest production plant, producing since 1902. We established a digital factory campus there, where we have a lot of robots and competences in data and AI. We invested a three-digit million euro amount. There we develop digital production processes rolled out to over 30 factories. Last week we introduced a new colleague, a humanoid robot named Apollo. We decided to invest in a very interesting startup from the US, Apptronik. We announced that and showed what we want to do. It's a Mercedes-branded humanoid robot. Let me give you a glimpse. Apollo, a super intelligent machine interacting with surroundings, helping workers with tasks. Built to navigate in human environments, it supports the team. Some jobs demand strength, precision, endurance. Apollo steps in where needed most, lifting, assembling, handling repetitive work. Apollo is designed to augment and support people, taking on heavy work so teams focus on what matters. So these three nice guys are helping us right now in Berlin. We introduced them in Hungary first and are rolling out. What do we want to do with them? First, quality inspection, because the combination of mechanics, physics, and AI can help on exterior and interior quality management. Also in intralogistics, especially where loads are heavy and tasks hazardous. We want to increase productivity and improve quality. It's very important that we get data and physics together. We use our MO360 ecosystem to train Apollos for automotive production. Making data available to everybody is the winning core topic for humanoid robots. We call it data democratization. We are coming out of silo organizations. With MO360, we make data available for everybody. We developed this system together with Siemens and Microsoft. Siemens provides an interface between equipment and cloud called MindSphere, and Microsoft Azure to store data and develop algorithms. More than 50,000 users in our production ecosystem use it. On one hand, we have 100 applications like digital shopfloor management, quality management. We develop that in Berlin and roll out globally. Not only in process but also in planning. Our most important partner is Nvidia. We create a digital twin of our factories to install and improve robot locations. We can refurbish production lines quickly and upgrade them for new models, like the CLA in Rastatt, Hungary, US, China, South Africa. Less change costs and higher speed. Of course, AI plays a big role. If you have data and devices like Apollos, what do you do with the data? For example, in the paint shop, we coded our own algorithm to increase energy efficiency by 20%. We are introducing chatbots for production so a worker can access data for standard process descriptions. Let me show a video on AI in our daily work. At Mercedes-Benz, AI is part of the team, designed to be intuitive as a chat. It helps make smarter, faster decisions. AI virtual agents make complex tasks easy, delivering real-time insights, optimizing workflows, giving everyone instant access to expert knowledge. No coding needed. Built on decades of production expertise, AI knows what matters. It learns, adapts, becomes smarter, becoming a natural part of production. This is just a glimpse of what we are doing with AI in our production team. The video was from Berlin. At the end, the combination of data, AI, and innovations like humanoid robots will drive innovations in production. I'm personally convinced, as we transformed car body work in 1970, that we will see humanoid robots. The question is with or without legs. That will be discussed in the next few years. We will see applications mainly in intralogistics, like picking parts to bring to the line, where AI enables productivity and reduces mispicks, and also in quality inspection. The physical shape has to be decided. What we are doing now with a real humanoid robot is working on the most sophisticated scenario, which can be scaled. I'm very excited that you have this conference. We need your innovation and scientific knowhow to develop these applications further. I ask you to exchange and discuss and bring robotics and AI to the next level. If you have ideas, contact me. We can apply them in our factories. Bringing a scientific idea to life in a real factory is the best feeling. Thank you very much.

So basically we have already tried it. We introduced it very soon in our running production. It was pretty okay in Hungary, and you have seen some interaction with team members, giving a star to a team member. What is very important besides all the data development is communication. This is very good. It also depends on the shape. A humanoid robot should not look like a soldier; it should look nice to be accepted. This should not be underestimated. If you combine everything, you get the acceptance level needed. I think it's pretty soon; you will see that in the next couple of years.

Obviously, occupational safety is extremely important in a production environment. We are focusing on that. We have a lot of experience with cobots and know-how on integrating them in a running assembly line. We are making progress. There are challenges, but what we can see, especially on the data side, we can copy some of the experience from the product side, like level 2++ in the car, to apply to robots or humanoid robots. This can be managed and controlled, but it's a focus area.

We are not only focusing on non-German or non-European partners. We are very interested in working with European and German partners. We have Siemens and Dürr as partners, very traditional German companies. However, we have the experience that with US and China-based startups, it's faster and less complicated. On the research side and startup side, why not use that benefit? On the other hand, we would be very happy to work with German and European startups as well, but there is a difference.

As always, if you work on technical challenges, the most important topic is competence and intelligence. I'm convinced that in Europe, at the startup level, in companies, and in the scientific environment, we have extremely high competence. The basis is there. Let's try to transfer that into reality, to focus on application. That is the most important topic. We always have a lack of speed and focus. When I work with American or Chinese companies, the application and focus are extremely high. But in Europe, you have extremely high scientific competence. So let's do that. We are the last who wouldn't work with partners from Europe or Germany.