🗓 EpiPad Uses the Forearm as a Trackpad

A 5×4 mm millimeter-wave radar on the watch strap reads tiny changes in reflected signals as fingers press the skin, leaving the display untouched.

EpiPad estimates pressure from reflection phase and uses lightweight machine learning to recognize gestures. The forearm supports scrolling, zooming, clicks, pressure levels, and multi-finger input.

In tests with 20 volunteers, it achieved 92.8% gesture accuracy and 90.2% pressure accuracy, including with people absent from the training data. Its roughly 10 mm distance accuracy beat touchscreen and tilt controls, with fewer errors and higher efficiency.

📊@tech