MEET THE FOUNDER

When accuracy alone is not enough

Meet Dr Zhenzhong Su, Co-Founder and CEO of Fixposition, building the navigation foundation that allows autonomous machines to operate reliably – even where satellite signals fail.

“I quickly realised that accuracy alone was not enough.”

Reliable navigation remains one of the biggest constraints on autonomy. Founded by Dr Zhenzhong Su, Lukas Meier, and Tonio Gsell, ETH Zurich spin-off Fixposition combines satellite positioning, computer vision, inertial sensing, and AI to help robots and machines understand where they are and what surrounds them. Today, its technology is used by more than 200 OEMs across agriculture, construction, logistics, mapping, and security, and the leadership team has been strengthened by Peter Mardaleichvili as COO and Ulf Claesson as Chairman. Zhenzhong discusses the technical and commercial insight behind the company and the challenge of turning European innovation into industrial scale.

What convinced you this was a company worth building?

A spark and a healthy dose of optimism.

During my PhD at ETH Zurich, I was researching centimetre-accurate RTK GNSS positioning, and in 2015 I built my first prototype. My colleague Tonio Gsell and I applied the technology to autonomous drones and vehicles, and it won the ESA Special Prize at the European Satellite Navigation Competition.

That recognition was the spark. It gave us the confidence to turn the research into a business and join ETH Zurich’s Pioneer Fellowship programme.

But as I began speaking with potential customers, I quickly realised that accuracy alone was not enough. While RTK GNSS delivers centimetre-accurate global positioning, industry needed precise positioning that would continue working everywhere - under trees, between buildings and wherever satellite signals became unreliable. GNSS on its own could never solve that problem.

The breakthrough came when we met Lukas Meier and Lorenz Meier, whose Pioneer Fellowship team was developing computer vision for drone navigation. We saw that deeply integrating GNSS with computer vision could bring precise positioning to environments where it had never previously been possible. In August 2017, the four of us founded Fixposition and entered the ESA Business Incubation Centre in Switzerland.

In hindsight, we were optimistic because we had not yet realised how hard it would be to turn a technology innovation into a scalable business.

What did you see in the market that others were missing, and how did that shape your focus?

The market was missing a precise, reliable positioning and sensing solution that works across both indoor and outdoor environments. That is essential for true robot autonomy on complex industrial, agricultural, and construction sites.

Some OEM customers initially compare our solution with an RTK GNSS component, but that is the wrong way to look at it. We provide a complete navigation stack – from 3D global positioning and orientation estimation to semantic and spatial sensing. Our algorithms handle the complex fusion of GNSS, inertial and camera data, delivering ready-to-use information for autonomous driving. So, our solution actually costs no more than buying all these sensor components and fusing them in-house, and we can deliver it faster and better, continuously improving performance and releasing new features.

In 2024, we made a deliberate decision to focus on solving difficult, high-value problems for industrial customers rather than pursuing the consumer robotics market. Customer volumes may be higher, but customers tend to prioritise cost over quality and performance, and those margins cannot sustain the growth of a Swiss team with relatively high talent costs. Industrial customers place greater value on the reliability and performance our technology provides, making that the right market for us to build in.

What is the biggest challenge to scaling from here?

The challenge we face, and one I believe many European startups share, is turning an innovation into an industrialised product and scaling it commercially at the same pace as companies in the United States and China.

Europe is often where technologies are invented, but not always where the commercial success is built. European startups have access to less venture capital than their US counterparts and less public funding than companies in China. Talent is also comparatively expensive – with the same one million dollars, a startup in China can probably achieve 1.5 to 2 times the output at equivalent quality. US engineers cost as much, if not more than Swiss ones, but US startups benefit particularly from the scale and speed of their capital market. They can draw on significantly more venture capital than their European peers and often raise considerably more during the early stages.

The good news is that AI is beginning to narrow the talent-cost gap. In our experience, a small team of excellent engineers working with AI agents can now achieve what previously required a much larger team. The capital and public-funding gaps remain, however, which is why deep-tech companies need more of the kind of support ESA provides.

What is the goal you are building towards, and what does the next stage of the journey look like?

Our goal is to provide the navigation foundation layer for Physical AI: precise global positioning and sensing that works everywhere, allowing robots and machines to operate autonomously, safely and affordably at scale.

Today, more than 200 OEMs, including Topcon, Bobcat, and Komatsu, use our positioning sensors on over 2,000 robots across agriculture, construction, logistics, mapping and security. Adoption is on track to double in 2026, and we have signed three mass-production supply agreements. We have taken Fixposition from zero to one, although it took slightly longer than we originally expected. We are now scaling from one to ten; the next major milestone is working out how to grow from ten to one hundred.

Together with our OEM customers, we are supporting more of their robots/machines moving from pilots into mass production and expect further volume and long-term supply agreements in 2027. At the same time, we are expanding from positioning into a complete navigation solution, including global sensing, foundation-model-based detection of driveable space and vision-based navigation that keeps drones on course under GNSS jamming and spoofing – everything needed as input for motion planning and control in an autonomous system.

Investment will allow us to expand our commercial and engineering teams, increase our capacity to support more OEM customers, and develop the navigation foundation model that can take Fixposition towards that next level of scale.