Anandghan Waghmare
Publications

Z-Band

IMWUT 2025Touch, swipe and pressure input on everyday surfaces from a wristband.

Z-Band is a wearable device that fuses bio-impedance and inertial sensing to enable robust, stateful touch detection on everyday, non-instrumented surfaces.

Stateful touch detection, achieved by injecting wide-band, swept-frequency RF signals through the wrist and analyzing their reflections and transmissions, enables sensitive touch detection even on challenging dielectric surfaces. Building on this, Z-Band supports a range of gestures, including swipes and taps, and senses touch pressure.

User studies show over 95% touch detection accuracy in stationary scenarios and over 90% in motion, without per-user calibration. The approach offers a seamless, natural way to interact with wearables and receive haptic feedback through touch gestures on any surface.

Z-Band, worn like a watch. It turns touches on ordinary, unmodified surfaces into input for your wearable.
Z-Band, worn like a watch. It turns touches on ordinary, unmodified surfaces into input for your wearable.
Tapping a refrigerator door: the band senses the moment the finger makes contact, even though the door has no sensors.
Tapping a refrigerator door: the band senses the moment the finger makes contact, even though the door has no sensors.
Your own palm becomes a touchpad: taps and swipes on the other hand are detected as input.
Your own palm becomes a touchpad: taps and swipes on the other hand are detected as input.
The underside of the band: two metal electrodes rest on the skin and send a tiny, harmless radio signal through the wrist and hand.
The underside of the band: two metal electrodes rest on the skin and send a tiny, harmless radio signal through the wrist and hand.
Inside the research prototype: a compact network analyzer measures how that signal changes the moment your finger touches something, alongside a motion sensor and a Raspberry Pi.
Inside the research prototype: a compact network analyzer measures how that signal changes the moment your finger touches something, alongside a motion sensor and a Raspberry Pi.
An early wrist-worn prototype we used to find which radio frequencies best reveal a touch.
An early wrist-worn prototype we used to find which radio frequencies best reveal a touch.