FMC³ Robotics unveils Doceo series to teach robots from workers

FMC³ Robotics is betting that the bottleneck for industrial robots is not algorithms but data. Its Doceo series is built around collecting information directly from factory operations while workers carry out their normal duties, rather than relying entirely on dedicated robot training facilities or simulated environments. Dr. Joachim Fetzer, president of Europe at FMC³ Robotics, frames the gap as one of qualified physical-world data that reflect how a task is actually performed, including errors and corrections.

The four components form what the company calls an industrial learning loop. RealCap gathers the raw signals, Hub synchronizes and annotates them, Cloud trains and tests models using data infrastructure from Vector Informatik, and Driver connects the results to different robotic platforms. Beyond the product, FMC³ is advancing two proof-of-concept projects — bin handling at a consumer electronics plant and electrode sheet handling at a battery cell facility — that will show how well human manufacturing experience translates into reliable robot operations.

The company is also working with Vector Informatik and PEM Motion on a planned German joint venture to standardize industrial data infrastructure, with the objective of making training data more consistent, traceable and reusable across factories and applications. If that works, the payoff FMC³ describes is less engineering and retraining when manufacturing tasks change.