The rise of vertical robots - a thesis on the service robotics market

We see a combination of technology enablers and market trends that push the path forward for service robotics startups.  Our colleague Marc Alexander Kühn dives deep into the latest trends and technological advancements that are driving the service robotics market forward and shares his personal view on the matter.

Introduction & Definition of service robots

Since my research time at the Oxford Robotics Institute, I have been closely monitoring progress on different fronts of the robotics and AI landscape. Starting from intelligent driving, over innovations for industrial robots, up to mega investment rounds in humanoid robots – a lot happened in this field over the last years. One area especially caught my interest and that’s service robotics: I believe now is the time that early-stage or soon-to-be founders can gain traction with respective products and services. In this blog post, I like to share my personal view on the space without any claim to completeness and look forward to further discussions. 

Let us begin with our UVC Partners definition of service robots: “A robot that performs useful tasks for humans or equipment excluding industrial and surgical applications. Service robots assist human beings, typically by performing dirty, dull, distant, dangerous, or repetitive jobs. They typically have a degree of autonomy.” This definition offers a clear distinction from industrial and surgical robotics, which are both more established.

Why now for service robotics

Diving deeper into the why now behind starting and financing a service robotics company in 2024, I like to highlight two separate dimensions: A) macroeconomic trends and B) technological enablers. On the macroeconomic side, there exist skilled labor shortages and rising minimum wages pulling for innovation in markets where physical human labor is essential. As a reference, 56% of all healthcare jobs in Germany are currently vacant. This illustrates a lack of qualified personnel for many service-related industries like hospitality (e.g. cleaning), care, inspection, or logistics. Also, minimum wages in 22 out of 27 EU member states are introduced with rising levels. For example, in Germany, the minimum wage increased within the last four years by over 30%. This makes it difficult for service-oriented businesses to hire enough personnel. One can deduct here that some kind of innovation is needed to address these challenges.

When we think about technological factors pushing for innovation in the service robotics sector, we can identify advancements in robotic cognition and more powerful computing units. Starting with the former, especially recent advancements in multimodal transformer models as well as novel reinforcement- and imitation learning methods allow faster training and better performance of robots. Another important part is the improvement of robot-human communication enabled by modern LLMs. They enable the translation of human speech to robot control instructions, which allows a natural way of communication that is essential for a high-performing service robot. On top of that, the tremendous improvement in compute unit power allows robots to run larger models locally. For reference, the Nvidia AI GPU compute power increased by a factor of 1000x in the last eight years. Moreover, this technology enables the processing of many input data streams like high-resolution video or lidar data simultaneously in real-time. Both developments are essential to advance service robots from a nice gimmick and research projects to be actually very helpful assets.

Those elaborations make a case for A) the increasing capabilities of robots that make service-like interactions possible and B) a strong market need for this technology. 

Segmentation of service robotics use cases

Based on the above reasoning on why now is an excellent time to engage in service robotics, I and our investment team further wrapped our minds around which use cases exist and what startups / scale-ups are active in the respective fields. You can find our findings as a non-exhaustive list in the figure above. On the one hand, we see vertical use-case-focused robotic solutions in spaces like logistics, inspection, healthcare, agriculture and hospitality. On the other hand, we also see hardware and software positionings that are cross-vertical like fleet management and robot operating systems. Moreover, there could be more established robotics companies entering the service robotics market from the industrial sector, like our portfolio company Fruitcore Robotics, that have all the necessary capabilities (from hardware development, software know-how, design capabilities to cost competitiveness). Nevertheless, I keep them out of this analysis because I don’t want to lose focus here.

Looking towards investment theses, there are of course, restrictions. More concretely, not all use cases are threatened equally strongly by the illustrated labor shortages. Also, we need to look for use cases that technically were not possible in the past and are now enabled by modern robotic cognition and improvements in computing. We also need to keep in mind that there will be use cases for which technology needs to advance further. On top of that, one needs to look at the production competitiveness compared to geographies like China and the price competitiveness compared to manual labor. So, finding the right vertical to engage in is not trivial.

Based on the above-mentioned restrictions and a large and growing market opportunity, we see large potential for early-stage startups in logistics, inspection, companionship, kitchen aid, and cleaning. Of course, this is just based on our assumptions, and we are always excited to hear the founders' perspectives.

Possible future service robotics scenarios

This leads us to the important question: What’s our takeaway for investing in service robotics startups? As always, it’s not that straightforward. I like to start by highlighting a few possible future scenarios.

Scenario 1: Robotic software capabilities are expensive to develop, which leads to an oligopoly structure of only a few companies developing robots and services.

In the first scenario, the assumption states that robotic cognition software skills will be a very defensible asset. This creates a barrier to entry for new players who like to create their own robots or robotic services. The knowledge about creating top-performing robotic cognition skills will exist only at a handful of companies, which then have the possibility to build a portfolio of robot hardware and robot services around the software core.

Scenario 2: Robotic hardware and software become commodities, and companies focus on delivering value in niches.

In this scenario, the opposite would be the case. Both hardware and software might become commoditized (e.g., through open-source AI models), which lowers the barrier to entry. This would lead to a scenario where every engineer might be able to build a service robot. The value of an emerging startup would then be mostly in winning distribution and usability in a specific vertical.

Scenario 3: Although strong complexity lies in robotic software development, an oligopoly of software companies might serve a large number of robotic niche application/ service providers.

Our third scenario is a hybrid of the first two. In this case, building advanced robotic cognition, including multi-modal LLMs for robot-human communication stays very complex and high-value. Nevertheless, complexity in delivering value in specific use-cases remains high through the need for strong distribution and usability. This opens up symbiotic relationships between software providers and robot service providers that buy access to the latest robotic software but still own the complexity of the vertical.

Although we at UVC Partners are techno-optimists, of course, we nevertheless want to flag scenarios that result in a negative sentiment.

Scenario 4: The performance of robots does not meet mass market requirements in the medium term.

This scenario is very important to keep in mind as this is at least partly the reason why service robotics did not already lift off 15 years ago. Although the technology pushes of the last few years mentioned above were tremendously impressive and important, most commercially relevant use cases are still in a pre-product-market-fit stage, which leaves a risk for insufficient performance.

Scenario 5: Regulatory and ethical concerns dictate boundaries of service robotics.

Another pessimistic scenario could be driven by regulatory constraints, especially regarding human safety certification, which hinders large-scale roll-out of robots, which interact in dynamic environments with humans. Also, ethical considerations about securing jobs and, therefore forbidding or limiting automation in certain areas might be a realistic fear – either driven by the government or by individual companies’ works councils.

Our investment thesis in service robotics

So let’s put all things said into perspective and try to aggregate investment theses.

Our interest in the service robotics market is driven by a base thesis stating that technological advancements in robotic cognition and compute units as well as macroeconomic pulls resulting from labour shortages and rising minimum wages make a strong case for a growing service robotics industry. Nevertheless, it’s important to state a remaining undeniable risk associated with the possibility of insufficient robot performance in the near future.

Based on my interactions with founders, customers, and robot experts, I believe in the future of vertical robots. My thesis is that we will have a world of 80%-95% commoditized hardware and software components in the near future, and the very valuable thing to do will be winning a specific use-case, meaning dominating distribution and fine-tuning a product that brings great usability to the specific vertical. This can either include business models that sell robots to sufficiently knowledgeable enterprises that know how to operate them on their own or could even mean offering robots as a service to replace traditional service personnel, e.g. in cleaning. Obviously, it is not yet clear whether the robotic cognition software stack indeed will become commoditized. Even if there will be an oligopoly of software suppliers, I believe that a strong need for vertical integrators and / or service providers exists to offer a strong and focused product offering to the customer.

In conclusion, it can be summarized that there are great opportunities in the service robotics industry. Founders who are looking for an emerging field with strong “why now” should look closely at opportunities in service robotics. And if you are already building a company in this area, I would love to have a chat with you.

Special thanks to my colleagues Nikolai Göbel and Benjamin Erhart, who worked with me on that deep dive, and to all the founders and investors I talked with in the last weeks.

Let us hear your thoughts!

I am excited to hear back from you regarding your take on the service robotics space! Please feel free to contact me at kuehn@uvcpartners.com – especially if you are a founder building in the space, regardless of whether your thesis aligns with mine or you see things differently.

About the author

‍Marc Alexander Kühn is an investment associate with UVC Partners’ Munich office. He holds a Master’s degree in Engineering and Informatics from the Technical University of Munich with a research stay at the University of Oxford and a Bachelor’s degree in Engineering and Management. His previous experience includes deep-tech venture capital, computer science research at Fraunhofer-Gesellschaft, technology strategy at a global corporation, and freelance business consulting.

Mail: marc.kuehn@uvcpartners.com

About UVC Partners

UVC Partners is a leading venture capital firm that invests in European B2B tech startups and has offices in Munich and Berlin. With more than €600 million in assets under management, the VC typically invests between €1 to €10 million initially and up to €30 million per startup in the areas of DeepTech, ClimateTech, Mobility, and Software/AI. As an independent partner of UnternehmerTUM, Europe's leading startup hub, UVC Partners has unique access to proprietary deal flow, more than 1,000 corporates and SMEs, as well as to talents from the Technical University of Munich, which belongs to the best European technical universities. UVC Partners' investment portfolio includes Flix, Isar Aerospace, planqc, Proxima Fusion, Reverion, Tacto, TWAICE, DeepDrive, STABL, and many more. All portfolio companies and founders benefit from the team's extensive investment and exit experience, their ability to build sustainable category leaders, and the network of UnternehmerTUM enabling them to speed up market entry.