I/O Braid is an interactive textile cord with embedded touch sensing and visual feedback capabilities. The technology transforms ordinary cords into smart interfaces that can detect proximity, touch, and twist gestures through a spiraling, repeating braiding topology. The same design allows for embedded fiber optic strands to provide co-located visual feedback. I/O Braid offers a unique solution for integrating interactivity into everyday objects like headphone cords and garment drawstrings without compromising their design, aesthetics, or primary function. Developed during my time at the Interaction Lab at Google.

Published at UIST 2018, where it received the "Best Demo" award.

My Contributions

ML Algorithm Development & Sensing Architecture

• Mutual-capacitance touch sensing architecture, leveraging a unique spiraling, repeating braiding topology with conductive yarns
• Learning algorithms & a training set for detecting multiple gesture types including proximity, touch, rotation, and pressure

Hardware Development

• Research prototypes for a variety of sensing architectures
• Data collection apparatus for training ML algorithms
• Standalone USB-C headphone demonstrator unit with on-device ML, acting as USB HID device and USB audio endpoint

Firmware Implementation

• Firmware for gesture detection pipeline, USB audio streaming, and HID control
• On-chip audio feedback for low-latency gesture earcons without USB roundtrip
• Deployment of on-device ML algorithms for detecting various gestures including taps, holds, grabs, and rotation
• Sensor baselining algorithms to improve reliability against drift, aging, and environmental changes