--- title: "Between Lab and Factory Floor: Where Are the Opportunities in Robotics? " date: 2026-10-28T14:00:00Z modified: 2026-10-28T10:18:49Z permalink: "https://www.htgf.de/en/where-are-the-opportunities-in-robotics/" type: post status: publish excerpt: "" wpid: 134240 categories: - Story language: - English post_translations: - pll_6ae0727a72cac featured_image: "https://www.htgf.de/wp-content/uploads/2026/10/Gernot_DLR-1-edited.png" timestamp: 2026-10-28T10:18:49Z tags: - Story - English - pll_6ae0727a72cac --- # **Between Lab and Factory Floor: Where Are the Opportunities in Robotics?** Dr Gernot Berger (HTGF) and Rüdiger Süß (DLR Startup Factory) on Physical AI, spin-offs and investment potential Robotics is at a turning point. Physical AI, significantly more computing power and a structural labour shortage are all converging at the same time. In mid-October, HTGF and the German Aerospace Center (DLR) jointly hosted [Robotics ](https://www.htgf.de/en/htgf-robotics-day-2026)[Day at DLR’s Oberpfaffenhofen site](https://www.htgf.de/en/htgf-robotics-day-from-research-to-real-world-applications/). Robotics experts from research met industry users and deep-tech investors. Now we’re following up: Where are the investment opportunities, and what determines the leap from the lab to the factory floor? We spoke with Dr Gernot Berger, Principal at HTGF, and Rüdiger Süß, Senior Investment Manager at the DLR Startup Factory. ![](https://www.htgf.de/wp-content/uploads/2026/10/Gernot_DLR-1-edited.png) _From left: Dr Gernot Berger (HTGF) and Rüdiger Süß (DLR). Photos: HTGF, DLR_ **Two weeks after Robotics Day: What insight or moment sharpened your view of robotics investments the most?** **Rüdiger Süß:** One question from the audience gave me particular food for thought: What do you actually need a five-fingered robotic hand for? The answer: Sometimes three or even two fingers are enough, and sometimes two hooks are the best solution. The same applies to humanoid robotics as a whole. It may be the hype of the moment, but it isn’t always the best solution. Humanoids make sense in direct contact with people, and when they share a workplace with humans – for example, when they are meant to carry out the same tasks with the same tools. From an investment perspective, this means: We won’t need as many humanoids as one might think. What is needed are automation solutions that deliver an economic advantage! So the reality check shows: Industrial robotics doesn’t necessarily need humanoids. It calls for flexible answers. **Gernot Berger:** What reinforced my view was how many industries now see concrete tasks for robotics. That was exactly our goal. Above all, we wanted to bring together users from industry, from large corporations to hidden champions. They were joined by researchers as well as some founding teams and investors. When users and researchers discuss things directly, it quickly becomes clear where a technology is really needed and what is still missing for deployment. For us, that’s an important complement to what we see in our deal flow. One example is medical robotics. In our deal flow, it is the second-largest application area after industry. During the tour of the DLR labs, we saw a very exciting robot-assisted solution for treating kidney stones. Applications like this show how far robotics now reaches beyond the factory floor. **Gernot, in your presentation you showed where founding momentum is emerging in HTGF’s deal flow – and where it isn’t. Where do you see the biggest gap between startup activity and market demand?** **Gernot Berger:** We don’t just look at where many teams are founding companies, but also at where there are still few. The gap is particularly striking in safety, certification and integration. That is the smallest segment in our robotics deal flow, even though these are precisely the issues that often hold back real-world deployment. The topic ran through the entire day. Without robust answers on safety, hardly any company will move a robot into routine operation. We also see comparatively few startups in sensors and actuators. Dr Werner Kraus, Head of the Automation and Robotics Research Division at Fraunhofer IPA, reported that China now has more than 150 manufacturers of humanoid robots but only a handful of component suppliers. That’s exciting for investors. Companies that supply key components can use their technology to enable many different robotics applications. ![](https://www.htgf.de/wp-content/uploads/2026/10/20261013-AB2_1017-1-1495x1080.jpg) _Dr Gernot Berger at Robotics Day 2026 at DLR’s Oberpfaffenhofen site. Photo: HTGF / Alexandra Beier_ **Rüdiger, you support DLR spin-offs and co-invest in them. Which robotics technologies are ready to make the leap into a startup, and how can you** **tell** **that a team is „investable“**? **Rüdiger Süß:** It’s not so much about maturity levels as about whether the application of the technology is clearly recognisable. But the most important question is: Does the team have all the skills needed to move a company forward? An investor needs to trust the people. Can they also handle taxes, legal matters, people management and customer acquisition? Can they manage money, think and act like entrepreneurs, and will they stay the course in tough times? And hopefully they are open to advice. Last but not least: How passionate is the team about what they do – or, as they’d say in football, do they have that drive towards goal? **AI-based robotics: hype or the next mass market? How do you see Physical AI as a driver of this development? And what does a startup need to show for you to believe in its potential?** **Gernot Berger:** Physical AI is opening up new possibilities for robotics. It is the fastest-growing segment in our robotics deal flow, albeit from a small base. What I find interesting is where the startups are emerging. Capital and attention are currently pouring into humanoid robots. But many new startups are working one level below that – on the intelligence that enables robots to learn. Dr Falk Kunadt from DLR described humanoids as a technology driver for robotics as a whole. Whether this becomes a mass market will be decided by the applications. Does the system work under real-world conditions? And what benefit does the customer get from it? We also look at whether the team can use data from its deployments to continuously improve its system. This data flywheel can create a lasting competitive advantage. **Rüdiger Süß:** Physical AI can be a game changer if it enables a „dumb“ machine to behave more intelligently – for example, by adapting to changing, unpredictable conditions. Ultimately, mass markets can grow out of that, too. In industrial robotics, I see it above all as an enabler. If AI speeds up how robots learn, and later, in operation, helps to analyse sensor data to predict developments such as wear and tear, then that has real value. What does a startup need to show? First, as I said, a clearly recognisable application. Second, genuine proprietary knowledge. Are they doing more than just combining other people’s knowledge and technology? Do they really have something of their own? If you own the core, you can’t be copied! That’s the decisive potential. For us, it has to be DLR technology, otherwise we don’t invest. ![](https://www.htgf.de/wp-content/uploads/2026/10/20261013-AB2_1368-1-1619x1080.jpg) _Impressions from Robotics Day. Photo: HTGF / Alexandra Beier_ **The path from pilot project to factory floor is the big hurdle. Which applications and business models do you think have the best chance of clearing it?** **Rüdiger Süß:** Automation will increase, but it doesn’t always have to be robotic. What matters is whether the startup’s solutions can be integrated quickly and easily into customers‘ existing processes. Companies don’t turn their production lines upside down overnight just because a startup comes along with a technically fascinating solution. They simply can’t. Only those who focus on customer value stand a chance of succeeding in the market. The same goes for business models. What suits the target customer, and how flexible and adaptable can I make my offering so that I reach as many customers as possible? **Gernot Berger:** In a pilot project, it is often enough for a system to work most of the time. In day-to-day operation, it has to work almost all of the time – and without an expert constantly standing next to it. My colleagues Viola Paszkiewicz and Alex Paß illustrated this very vividly in their talk together with Pierre Bürkle (Schneider Electric). That’s why applications where errors are manageable and the system improves with every deployment have good prospects. In terms of business models, Robotics-as-a-Service is an interesting approach. The customer pays for a service instead of buying the machine itself. That can make it easier to get started, because there is no high upfront investment. But the model has to work for the startup, too. It has to finance the systems and operate them reliably. So we look closely at how much customisation each new customer requires. If every project is a one-off, the business is hard to scale. **Where do you see the greatest investment potential in robotics over the next five years? And what advice would you give founding teams starting out now?** **Gernot Berger:** I find Physical AI infrastructure particularly exciting – that is, technologies that connect AI models with hardware and make them usable in real-world operation. Add to that solutions for safe deployment and applications in areas facing labour shortages. Another area we are watching closely is industrial AI. By that we mean AI that works directly within industrial processes, for example in quality inspection, predictive maintenance or production planning. It isn’t always robotics in the narrower sense, but it is closely related. Initial analyses of our deal flow suggest that this area is also growing strongly. My advice to founding teams is to start with a specific customer problem and get into the field early. Bring in the right partners from research and industry. And keep an eye on what can be transferred to other customers. That’s how a good solution becomes a scalable company. **Rüdiger Süß:** I think the greatest potential lies less in highly specialised parts and more in those needed in every robot, such as chips, combined with customisable software solutions, or actuators. Anyone who can supply reliably and at high quality in this field will be among the long-term winners. We’re seeing a very good example of this right now with NVIDIA. Its chips are used in many different robotics solutions. And my advice to founding teams: Have the courage to move fast. Seize the opportunities that are opening up now. And: Think big! Thank you both for the conversation and your insights. **About the interviewees** **Rüdiger Süß** is Senior Investment Manager at the DLR\_Startup Factory within DLR’s Innovation, Technology Transfer and Scientific Infrastructure division. He is responsible for co-investments in DLR spin-offs, matchmaking with investors, support to increase investment readiness and investment attractiveness, and venture clienting. **Dr Gernot Berger** is a Principal at HTGF in the Industrial, Climate & Deep Tech investment team and has been with HTGF since 2019. A physicist specialising in photonics, quantum technologies and energy, he himself helped take a photonics startup all the way from technology to exit.