Natural User Interface control of a Parrot drone using a wearable piezoresistive fiber sensor — a novel conductive composite developed at Empa. Gesture recognition from skin-surface strain data mapped to drone flight commands.

As robotics and autonomous systems enter everyday environments, the question of how humans interact with them naturally remains unsolved. The client needed a working prototype demonstrating NUI control of a drone — using body motion alone, without any handheld controller.
At the heart of the system was a smart strap wearable sensor based on a novel conductive composite developed at Empa's Laboratory for High Performance Ceramics. The sensor — nano-textured carbon particles in a thermoplastic elastomer — produces a highly elastic piezoresistive fiber that monitors surface strain at the tendons, detecting hand open/close motions that IMU sensors cannot capture.
We built the gesture recognition pipeline on an Adafruit Feather HUZZAH (ESP8266) with custom ADC integration, then mapped gestures to drone commands via the Parrot SDK using Node.js — chosen specifically because it allowed rapid iteration between hardware testing and SDK integration.
A functional prototype demonstrating real-time gesture-to-drone control via a wearable sensor. The underlying sensor technology has since been developed further for vital sign monitoring under the carunda24.com product line. The project demonstrated idezo's ability to bridge novel materials science with consumer robotics in a working prototyping timeframe.
STBL Medical Research AG / Empa
2017
Have a similar challenge?
Talk to us