How Switzerland became the world’s most deeptech-focused startup economy
ETH Zurich's historic main building (HG). The university plays a crucial role in Switzerland's deeptech ecosystem (photo: ETH Zurich)
When Gravis Robotics raised $200 million from SoftBank in August, the size of the deal attracted almost as much attention as its autonomous construction technology.
Founded in 2022, the company described the investment as the largest Series A in construction robotics. It valued Gravis at $1 billion, making it the fifth ETH Zurich spinout—and the university’s first robotics company—to achieve unicorn status.
As MTN reported, Gravis develops physical AI that enables excavators and other heavy machinery to reshape their surroundings autonomously. Its systems combine cameras, sensors, three-dimensional mapping and machine data to interpret changing terrain and control the equipment’s hydraulics. Existing construction machinery can be retrofitted rather than replaced.
Gravis looks exceptional. In Switzerland, it is also part of a pattern.
The Swiss Deep Tech Report 2026 finds that 63% of all venture capital invested in the country between 2020 and 2026 went into deeptech. That is the highest share recorded in the report, ahead of China at 56% and the United States at 54%.
The figures point towards an unusually concentrated startup economy. Switzerland’s deeptech advantage rests on technical universities, specialist industries, international talent and dense networks connecting research with capital. As investment moves towards artificial intelligence, robotics, advanced computing and biotechnology, that structure is becoming particularly valuable.
ETH Zurich and EPFL built the deeptech pipeline
At the centre of the Swiss deeptech ecosystem are ETH Zurich and the Swiss Federal Institute of Technology Lausanne, better known as EPFL.
The report ranks them first and second among European universities for new venture-backed deeptech spinouts. Their contribution goes beyond licensing patents. Laboratories, technical talent, entrepreneurial programmes and connections to industry allow research teams to develop companies around technologies that may require years of further work before reaching the market.
Gravis emerged from ETH Zurich’s Robotic Systems Lab after almost a decade of research into autonomous heavy machinery. The same laboratory produced ANYbotics, whose four-legged robots inspect industrial facilities, alongside a wider group of robotics spinouts working on areas ranging from security and disaster response to autonomous transport.
ETH Zurich’s Department of Mechanical and Process Engineering alone had produced 120 spinouts between 1986 and the third quarter of 2025, approximately 20% of all ETH spinouts.
This creates a compounding effect. Students encounter researchers who have commercialised technology before. New founders can recruit people with both scientific and startup experience. Investors become more familiar with long development periods and technical risk. Successful companies then provide experienced employees, advisers and potential founders for the next generation.
Universities in many countries produce strong research. Switzerland has become particularly effective at surrounding that research with the people and institutions needed to turn it into a company.
Industrial expertise prepared the ground
Swiss deeptech did not emerge from universities alone. The country already had specialist knowledge in pharmaceuticals, precision engineering, medical devices, industrial automation, microelectronics and sensors.
Those industries provide experienced workers, suppliers and potential customers. They also give startups practical problems to solve.
Switzerland’s biotechnology base remains the largest part of its venture-backed deeptech ecosystem and has created the greatest share of its enterprise value, according to the report. Companies are now combining that life-sciences heritage with AI, robotics and advanced computing.
The composition of new company formation is changing accordingly. AI and machine learning represent one in four Swiss deeptech companies founded since 2022, up from 11% during the 2010–2021 period. Robotics has also expanded its share of new company creation.
Since 2020, Switzerland has produced 3.5 times more venture-backed robotics startups per resident than the US, according to the report.
In technologies classified as the future of computing, Switzerland recorded seven times more European patent applications per resident than the EU average. The country’s established capabilities in microelectronics, sensors and thermal engineering now support companies working on problems such as high-speed chip connections and cooling increasingly powerful processors.
New computing and robotics companies are therefore being added to capabilities accumulated through pharmaceuticals, manufacturing and precision engineering rather than replacing them.
A small country with a dense technology network
Scale can be an advantage in venture capital, but Switzerland shows what density can achieve.
The country had a permanent resident population of approximately 9.1 million at the end of 2025, with much of its deeptech activity concentrated around Zurich, Lausanne and Basel. Founders, researchers, investors and established companies operate within relatively close networks.
Switzerland also sits within what the report calls the Alpine technology cluster, connecting concentrations of research and industry around Zurich, Basel, Munich and Milan.
Its technical universities draw researchers from around the world. The report estimates that Switzerland has the world’s highest concentration of AI researchers relative to its population, approximately twice that of the UK and US.
This concentration comes with high salaries and operating costs. For software companies that can be built almost anywhere, Switzerland may be an expensive base. For startups depending on scarce scientific talent, specialised laboratories or access to pharmaceutical and engineering expertise, proximity to those resources can offset some of the additional expense.
Swiss deeptech companies raised a record $2.6 billion in 2025, roughly five times the amount raised in 2015. Across the 2020–2026 period covered by the report, the country attracted $1,470 in deeptech venture investment per resident. That placed Switzerland first in Europe and among the top three countries worldwide, alongside Israel and the US.
Switzerland’s late-stage funding gap
The Gravis Robotics investment also exposes the principal weakness in the Swiss model: its main investor SoftBank is Japanese, not Swiss.
International investment demonstrates that Swiss startups can compete for global capital. It also means that a substantial share of the investment returns from successful companies may flow to shareholders outside Switzerland.
Foreign investors provided 88% of the capital in Swiss deeptech rounds above $100 million between 2023 and 2025, compared with 75% across Europe. US investors alone supplied more than half of Swiss late-stage capital.
Swiss investors remain more prominent near the beginning of the journey. They provided approximately 36% of the capital in smaller rounds of up to $15 million, according to the report, but their share fell to 12% in rounds above $100 million.
Switzerland has therefore built an effective research and company-creation system without developing a comparable domestic pool of growth capital.
Deep Tech Nation Switzerland wants to mobilise CHF50 billion in venture capital and create 100,000 jobs by 2033. Persuading pension funds and other institutional investors to finance high-risk technology companies over long periods will be one of the harder parts of that ambition.
Foreign investment can give startups the capital to scale while remaining connected to their Swiss research and engineering base. Greater domestic participation would allow Swiss institutions to capture more of the financial returns if those companies succeed.
What other countries can learn from Switzerland
Switzerland’s deeptech model cannot be reproduced by announcing a fund or attempting to create a technology cluster from scratch. Its strongest advantages were built over decades.
Other countries can, however, copy parts of the system.
The first is sustained investment in technical universities combined with practical routes for researchers to become founders. Academic excellence produces greater economic value when technology-transfer teams, early funding, laboratories and experienced advisers remain available after the initial discovery.
The second is specialisation. Switzerland did not attempt to lead every area of technology. It built from existing strengths in life sciences, precision engineering, automation and sensing. AI and robotics are now extending those capabilities rather than displacing them.
The third is density. Connecting universities, established manufacturers, investors and public institutions within recognisable clusters can be more productive than distributing support thinly across unrelated projects and locations.
Finally, early-stage support must be connected to late-stage capital. Europe produces many technically strong startups but often depends on foreign investors when they require the much larger sums needed for factories, clinical trials or international expansion.
Switzerland demonstrates both the benefits of an effective research-to-startup pipeline and the limits of leaving its final stage incomplete.
Gravis Robotics embodies the result. Research conducted inside a Swiss university became a company applying AI to one of the world’s least automated industries, then attracted what the company describes as the largest Series A in construction robotics.
The question facing Switzerland is no longer whether it can produce deeptech companies. It is whether the country can finance more of them through industrial scale—and retain a greater share of the value its laboratories create.
Liked this article? You can support our independent journalism via our page on Buy Me a Coffee. It helps keep MoveTheNeedle.news focused on depth, not clicks.