In a paint booth at ENGEL Austria's Schwertberg factory, a worker picks up a handheld device studded with sensors, moves through a spray pattern the way she has done thousands of times before, and sets it down. Sixty seconds later, a robot arm is running the same motion on its own. No code was written. No robot programmer was in the building.
That is RoboTwin, a Prague startup founded in 2021 by three Czech engineers. The company has built a no-code platform that captures a worker's physical demonstration of a manufacturing task and translates it into instructions for industrial robots. The device records the worker's motion and generates the robot program without requiring a specialist programmer.
"We started with jobs that are ugly, dirty and unhealthy for workers to do manually," Megi Mejdrechová, RoboTwin's co-founder and chief technology officer, told the EU's Horizon research magazine. Mejdrechová, a mechanical engineer trained at the Czech Technical University in Prague, was named to Forbes Czechia's 30 Under 30 list in 2025. "The robot basically copies the human demonstration," she said. "People with no coding skills can transfer their know-how and experience to robots."
The co-founders are Mejdrechová leading technical development, Ladislav Dvořák as chief executive, and David Polák handling sales. RoboTwin works with robot arms from FANUC, ABB, and YASKAWA, three of the largest industrial robot manufacturers globally, targeting surface treatment tasks: spray painting, powder coating, grinding, sanding, and polishing. These are jobs that are physically demanding, requiring workers to hold heavy equipment in awkward positions repeatedly, which makes them both expensive to staff and prone to injury.
ENGEL deployed RoboTwin across four robots running two paint booths at its Schwertberg facility. Viktor Hadač, ENGEL's head of painting and welding, told Startup Kitchen that the company robotized the booths without a dedicated programmer. "We had been searching for a way to automate our paint shop, and no solution on the market met our needs," he said. "Now we can robotize our two painting booths with four robots without a single programmer."
GalateK, a Czech surface treatment company, said it generates robot programs 66 percent faster using RoboTwin compared to conventional methods, according to a testimonial on RoboTwin's website. That performance claim has not been independently verified.
The traditional alternative to RoboTwin's system can take days of specialist programming work, and robot programmers are expensive and in short supply.
RoboTwin raised 1.25 million euros in pre-seed funding from BD Partners, Y Soft Ventures, EIT Manufacturing, and business angel Jörg Rupp. In 2025, it became the first Czech company in three years to win backing from the European Innovation Council (EIC) Accelerator, securing a package worth 4.7 million euros: a 2.2 million euro grant plus 2.5 million euros in co-investment. The accelerator accepted 40 companies out of 959 applicants that cycle, a 4 percent success rate. RoboTwin's team has 17 employees.
The labor context here runs in both directions. Surface treatment jobs are genuinely unpleasant: workers wear protective equipment in poorly ventilated booths and face exposure to chemical coatings. Labor shortages in these trades are real, as Dvořák told EIT Manufacturing. "People are often not happy doing these things and there is a lack of workers willing to take these jobs." Automation fills a gap that workers have already abandoned.
The historical record is less comfortable. Research by the National Bureau of Economic Research found that the diffusion of computerized numerical control (CNC) machine tools in U.S. metal manufacturing during the 1970s and 1980s caused declining employment for less-educated production-floor workers even as it raised demand for computer-literate roles. The EU's Horizon research magazine noted that the automotive industry leads robot adoption in Europe, with about 23,000 new robots added to production lines in 2024 (Europe only). The global figure for industrial robot installations was 542,000 units in 2024, according to IFR World Robotics 2025. Those robots replaced workers.
RoboTwin's next phase, funded by the EIC package, pushes toward what the company calls AI-driven program generation: showing the robot the part and letting it infer the correct surface treatment motion without a human demonstrating it first. That would shift the product from a worker-enabling tool toward something closer to full task automation. The company has partnered with Wandelbots using the Wandelbots NOVA integration for robot-agnostic programming.
Every published account of these deployments traces back to the same Horizon Magazine piece published by the EU's research-and-innovation service, and to RoboTwin's own website and materials. No independent reporter has visited one of these factories to see the system running in person. The EIC grant is a matter of public record. The customer testimonials are specific and attributable. The funding numbers are documented. But claims about deployment scale or productivity impact have not been independently audited.
A faster way to program a robot does not decide who benefits from the automation. RoboTwin's founders speak consistently about empowerment and accessibility. What happens to the worker after the robot learns the job is a question the press releases do not address, and it is the one that matters most.
† Add footnote: 'Source-reported; not independently verified.'