Follow the Gradient Why the Last Mile Breaks Most Robotics Startups, with Roland Siegwart, Professor at ETH Zurich
Jan 15, 2026
Roland Siegwart is a Professor of Autonomous Systems at ETH Zurich and a key figure in Europe's robotics landscape. He discusses the critical challenge of moving from prototype to market-ready systems, emphasizing the often-overlooked last ten percent of development. Roland highlights the importance of early revenue from practical applications and the necessity for deep tech founders to remain patient while navigating long timelines. He also explores the balance of ambition with survival, the need for business-minded collaboration, and the urgent sectors where robotics can make a significant impact.
AI Snips
Chapters
Transcript
Episode notes
Long Runways Hide Real Market Pressure
- Deep tech startups face very long journeys where early capital can create dangerous comfort.
- Roland Siegwart warns many teams realize later they need paying clients, not just awards or funding.
Get Early Revenue With Practical Use Cases
- Seek early, realistic revenue by selling simpler, valuable applications rather than chasing the final vision.
- Use those early wins to fund next steps and make long-term scaling credible for investors.
The Last 10% Is Often The Longest
- The jump from a lab prototype to market-ready product typically costs about ten times more effort.
- Roland emphasizes the last 10% often takes as long or longer than the first 90% to finish.