Bosch-Geschäftsführerin Rückert: "Bei allen größeren technologischen Umwälzungen haben sich Arbei...

Many voices say that Germany is not leading in artificial intelligence, but if you look at the area of industrial artificial intelligence, which is a sub-area, we are very much at the forefront because we are simply very good at developing complex physical products, machines and products interplay, and we can also create ecosystems of science, startups, and business. So I see a great opportunity to advance industrial manufacturing, strongly supported by AI, in Germany and also in Europe.

Larissa, thank you for the invitation.

Yes, AI is a key technology for Bosch, an innovation booster. That means we naturally want to offer our customers the greatest possible added value. The moment I add software or artificial intelligence, I have a completely different lever. We have that in our products in mobility, but also in the household – we have great examples in the kitchen and also in manufacturing. Yes, the products also have to be manufactured, and there you can do a lot with artificial intelligence.

Well, at Bosch we already have 44,000 software developers today, by mid-decade 50,000. And you can imagine many of them are also involved in AI. We have a central research center, but also in all business areas. And there is a typical thing: great people attract great people. And I have to say, at Bosch we have many talents with AI expertise who also have connections to universities and thus attract more. It is a topic that we also have to ensure that in education and training, in schools and universities, these topics of artificial intelligence, but also the new models towards generative AI, are taught and trained. Absolutely, but the most important thing is that we give children an interest in learning and also drive creativity, because that is what we need in the future.

But we are very proud of our tradition. We have been working with very high-quality products for a long time, also in hardware. But what I find even more fascinating is that we are now at home in both hardware and software, and it is precisely this combination that makes the difference. Yes, so when someone at our company programs software, they don't just see it on the screen – then the car brakes, or the pump in the factory pumps, or the oven stops because the pizza is ready. And that is simply incredibly fascinating for software developers, but not only for them, also for our hardware colleagues. They think, 'Wow, I have provided even greater added value for our customers and consumers.' That means the DNA and also the culture will change. Yes, of course, cultural change is the basis for the entire digitalization, but it is a huge opportunity, and our employees are practically demanding it. We have many great, intelligent people who say, 'Let's go into the transformation because we want to offer the customer even more.'

Well, I did have a chemistry set, but the interest was that I always liked to take things apart and try to put them back together, and I actually thought I could predict how the different elements would combine. As a child, we had a radio – back then – I took it apart and tried to put it back together. The success was not completely overwhelming; it didn't work afterwards, but I learned a lot. And in chemistry, it's similar: I bring different substances together and at the end I try to predict what the result will be. It's actually quite similar in software – there I have certain software commands or logics that are processed and an expected result. In AI, it's about probabilities, so it all has a very similar systematics.

No, there wasn't. It came from my own great interest in the possibility of researching something completely new, but also in a certain way with analytics and influencing things in advance. And that great interest was enough to get me through to my doctorate.

Yes, when I started studying chemistry, the interest was more in research and developing completely new things for the environment or for medicine. But over time, I realized that it is very important for me to interact with people, and I didn't want to spend my whole life alone in the lab. That's something you learn during your doctorate. So alongside my hobby of software – or 'Kobold' as it was called back then during my studies – I also increasingly dealt with a basic study of business administration and broadened myself. And then my first station was directly at SAP, where I stayed for 20 years.

Yes, when you look back sometimes, Larissa, a common thread appears. And I have to admit, I didn't see that thread at 18 or 25. But what I realized is that it's actually a combination of domain knowledge combined with software and now more recently AI and people. And that's what brought me to SAP. It was a very fascinating, young, cool company at the time, and they were looking for people who spoke the technical language of chemistry and pharmaceutical companies – domain knowledge – and combined that with software. I already had a penchant for that and some basic knowledge, and that was a wonderful entry for me, working in a team and with customers. And there's nothing as great as positive feedback from customers or colleagues.

Well, you start by developing software yourself, then you deal with the software development process and so on. But what is formative is also driving new business areas and new business models. And the new business models are, for example, the topic of digital supply chain – how can I intelligently connect manufacturing, logistics, transportation, combined with planning, using certain algorithms to know what inventory I need where. And then the whole area of the Internet of Things was a completely new area for SAP at the time, and pushing into that because it's not a typical software area – SAP doesn't have things. And that was particularly formative, being responsible for products and solutions in digital supply chains and IoT.

Well, Silicon Valley is not just a half; it's really a great experience. Anyone who has the chance to spend some time there should do it. It starts with having an incredible number of partner companies practically in the neighborhood – you get to know them, you talk briefly with Google, with Meta, or with Nest, or whatever – many companies. Secondly, it's this topic of taking a risk and possibly not making all your topics successful. That's not just said; it's really part of the culture. And I learned a lot during that time and hope I brought some of it back. Another thing is that our children were in kindergarten and school at that time, and just through the parents you get a network into other companies, other ideas. It's very enriching. And what's also great, what I personally find great, is that the sun shines a lot – you can dispute that, but when you wake up in the morning and look out and the sun is shining, it's a bit easier to be creative and innovative. If it's raining or foggy from the window, at least it's a very nice effect.

Yes, AI was already very relevant in the area where I ended up, with digital supply chains and the Internet of Things. The topic of predictive maintenance was discussed then and is still discussed today. And of course, when you can exchange ideas with colleagues from other companies or consider whether you might integrate a map – a navigation map – into your own system, such partnerships can arise or have arisen.

Well, first of all, you have great associations, and we have even more in our portfolio. The thing is that we are currently developing exactly into a hardware and software player, and that is naturally a great advantage when you have both in one house. And now with an area of 44,000 software developers, you can imagine that it is also important to build some things internally. Because we naturally have a lot of software, for example in the automotive control units, because we've been at it for a long time, but also software for analytics in the cloud or other topics where we also strengthen ourselves externally and perhaps bring in one or another complementary capability. And so I was very happy to have the chance to strengthen Bosch here.

I don't know, Larissa, at Bosch – and that was one of the reasons I decided to come to Bosch – they drove the topic of the Internet of Things and also artificial intelligence very early on. A separate center for artificial intelligence was founded as early as 2017, and the networking of devices was started early on, anchoring it in more and more parts of the portfolio, not just in washing machines. So yes, it fit well, but a really good foundation was laid. And a foundation-owned company may also have an advantage – not just words but also deeds. Investment really flowed and was built up.

I'll give a few examples right away because I think that's always the easiest way to understand. But the important thing is that we set ourselves the goal for 2025, but we will achieve it by the end of this year – that's two years faster. And why do I emphasize that? Because we are often very technology-oriented and want to do everything broadly, but when it comes to implementation speed – and I'm speaking for many companies here, maybe a bit ahead of our location – it can sometimes be a bit faster, to put it that way. And that's why it was very important to us to bring it into products and manufacturing very quickly, not just in research, where of course very good work is also done. So how can you imagine it? Let's go into manufacturing. Now let's go together into our semiconductor plant in Dresden. Yes, that is – let me introduce it: data is collected from the production systems, from the products, from the environment. We collect per second – we calculated it – 500 pages of paper could be described, or 42 million lines of text per day. You can imagine that, or maybe not. Why is artificial intelligence helpful there? Because a human would need years to evaluate it. And we build a digital twin of the manufacturing where you can also see how the machines are and how the products flow through, and then we can – if there is a process deviation, anyone in manufacturing knows you don't want to stop live production, that always has a big impact. Now in this digital twin of manufacturing, you can test what you would do to make the process run smoothly again, or an anomaly is detected, or I want to change the sequence, and then practically live at the moment when I have tried it on the production process – that already works today. If I go in the direction of a few products – we always think in our guiding principle 'Technology for Life'. That means you will now hear about products and solutions that improve life, make it easier, save lives, and also try to mitigate climate change. Now we need concrete examples. I'll start with a life-saving example: fire detection. Yes, so if you have a high room – we have fire detectors that are also very intelligent, but we also have fire video for fire detection. That means you also have the image and can then detect fire much faster in high rooms – you can imagine a very high ceiling, the smoke takes forever to rise. But if I have a visual image there, I can recognize it much faster, multiple times faster, and thereby of course detect fire very early in an airplane hangar or in large production halls, and thus prevent larger fires. And it can also distinguish between a Christmas candle and fire – sometimes very important, you can imagine. But when it comes to fire detection and saving lives, we brought out the first gas sensor with artificial intelligence last year. And why is it so great? We work with a company called Dryad that does wildfire detection. And now you probably know that because I'm also involved with the climate topic, but wildfires have an enormous negative impact on CO2 – it's almost as much as the entire transportation system, you can't imagine, but it's incredible. And this gas sensor – you attach it to various trees with this partner company – and it can detect very early, even under the moss layers or leaves, a smoldering fire because it uses artificial intelligence to recognize the gas combination that indicates a wildfire is starting. And if you can prevent just one large wildfire, then it's already more than worth it. Or if I now go into mobility – what is also important here, for example, in emergency situations to brake quickly or also to counter-steer so that you have to do less. And we have a software combination that mediates strongly between brake, steering, and also the drive, and thereby naturally has the possibility to significantly shorten the braking distance. ESP is already one of Bosch's inventions, and this builds on that. But we are talking about digital driving assistants. Yes, exactly. But just the idea that I connect brake, steering, and drive, and also connect that with software, and thereby I have a shorter braking distance on the one hand, but on the other hand I also have to counter-steer less. And that is exactly – yes, this detection of where I am with the brake and steering, and then finding the optimum. These are topics where you also connect very strongly via software. And then I go into the household. Yes, so if the topic is I want to bake a cake, or I come back from a soccer tournament with my son or with others in the car and want a pizza – the pizza should ideally be ready when we get home. And our intelligent oven can actually preset the browning levels. You set the browning level once, and then the oven knows exactly when to stop. Or I have – how does the pizza get into the oven? Yes, you caught me there. That has to be done by my husband. Okay, but you don't trust him to find the right settings, so you check from the car. I see absolutely – you also have to see, nobody wants to stand there for 20 minutes. What the system can also do is turn off when the cake is ready. I've unfortunately had that happen often – you come home late in the evening, the next day is the child's birthday (now they are teenagers), and you have the cake in the oven and neither my husband nor I actually want to stay up until the cake is done because we're already too late. And then the intelligent oven can also help.

Well, firstly, it is by no means an end in itself. You always have to start from the customer's perspective – customer problems, customer added value, consumer added value. If you say that topics don't help you in the kitchen, then you don't need them. But there are simply many topics like safety in the car or where I said I can avoid throwing away food that in my eyes are very relevant. We can also look at a few energy efficiency topics in manufacturing, for example, or in the household. That means for us, it is a key technology to give our customers and consumers added value, but also for us to drive business and efficiency. I already told the example of the factory in Dresden. We also have an analytics with AI that detects anomalies in all 2000 production lines and thereby brings an efficiency increase. Yes, that means for us, digitalization just like artificial intelligence is business. And that is a very important statement – that it is not technology for its own sake.

You would still see a considerable number of people, because it's also about the interplay – the digital factory, the factory of the future – and the people who, for example, now think about how to bring in new variants, who also have a certain control system of the process, who see the digital twin in front of them. I have also been there and I warmly invite you to come and see it together. And when such process deviations occur, they intervene correctively. In my eyes, it is a very good interplay between humans, machines, and in this case also artificial intelligence and software, where everyone has their place. So the human does more creative, forward-thinking work; the machine – there is also robotics, yes, which does heavier work; artificial intelligence makes the digital twin because it can evaluate the many data that humans cannot. And thereby it is, I would say, a mutually supportive togetherness.

With all major technological upheavals, jobs have shifted. For example, gas lamplighters are gone. Similarly, when robotics began, there were concerns, but at Bosch the number of employees has increased even with AI. We are now over 420,000 employees. New job profiles have emerged. We have an extensive AI training program. New profiles like data scientists didn't exist ten years ago. We must also consider the demographic shift: half of Germans are over 45, one-fifth over 66. AI can help with the skills shortage. In planning, we account for these shifts.

The first good thing is that in many cases our people drive the change themselves. They are interested in training. We support that. We spent over 260 million euros on training last year. In 2019, we started an AI training program with basic, advanced, and expert levels. We aimed for 30,000; we are already over 26,500. Additionally, each business area has targeted training. We collaborate with universities, and we have Bosch Tube with over 20,000-30,000 video formats on digital and AI topics. So training is a top priority.

It varies. There is basic knowledge that covers the opportunities of AI. AI is helpful. ChatGPT has allowed people to interact with AI and see it's not so bad. There is also deep technological know-how: different models, algorithms. We focus more on business opportunities, productivity, with examples. For instance, in a Reutlingen factory, a production planner was skeptical but later called the AI suggestions genius. So we start with a general training and then tailor it to specific areas like home, building, mobility, and to the audience, whether managers or workers.

Yes, that's why I say training and communication are key. We communicate the opportunities of AI through various channels that reach all employees. We set up a format called 'If Bosch knew what Bosch knows' where colleagues share successes and lessons learned in digitization and AI. This fosters cultural change by sharing real experiences, including mistakes.

Yes, we have a research area concentrating AI experts, service units that think about industrializing AI, and experts in each business area. So there is a concentration, a translation unit to spread it, and then local implementation. It's not centralized—that would be outdated. It must be international and across all businesses.

We have close university collaborations. But key is that good people attract good people. When they tell peers about what they can do at Bosch—writing code that makes a car brake or connects a photovoltaic system with a heat pump, increasing renewable energy use by up to 50%—that's very attractive. We are international, and we offer varied career paths: you can start as a developer, then move to a corporate startup, direct customer work, or research. That variety is attractive.

We will develop AI globally. We already have research centers and business areas worldwide. The key is not where it's developed but how. We follow common ethical rules: the AI Codex, based on Bosch values, established in 2019 before any EU regulation. All Bosch developers, regardless of location, follow the same principle: 'Technology for Life.'

To increase acceptance of AI, we need trust. Basic regulation helps build that trust, as long as it's not over-regulation. We welcome the EU AI Act and are engaged in the discourse, especially on generative AI. We also work with other companies in the Digital Trust Forum to develop an AI trust label, similar to energy efficiency labels, with criteria like fairness, data protection, security. This label would help customers quickly see if a product meets standards.

Many say Germany is not leading in AI, but in industrial AI, Germany is at the forefront because we are strong in complex physical products, machines, and ecosystems. I see a great opportunity to strengthen industrial manufacturing with AI in Germany and Europe. The question is less where we do it but how. We should focus on our strengths: AI in manufacturing, which may shift value chains. That's a good direction.

The opportunity is to drive cultural change through the employees. Our people want it. We support it with communication, training, career paths. We also ensure software and hardware colleagues collaborate effectively; we wrote a Software Manifest to clarify concepts like APIs—a colleague from manufacturing thought it was a standardized plug. Cultural change means offering opportunities, respectful interaction, and translation between disciplines. Additionally, generative AI is democratizing access: you can interact in natural language with systems, making it easier and more approachable.

They should study what they enjoy. You spend so much time at work; it's important to look forward to it. Creativity will become more important. The most important thing is that they have fun and find fulfillment, perhaps in climate, health, or new technologies. Also, as you can see from me, you can always switch to a software career later.

Absolutely, very welcome. It was a lot of fun, thank you, Larissa.