2026 · MIT Physical AI Hackathon
Textile Engineering · Electrical Engineering
Physical AI · Affective Computing
Role: Textile Design & Fabrication, Hardware Integration, Affective Inference Pipeline
A knitted biosensing wearable converting sender's voice messages into receiver's tactile feedback.
ConnectQ is an embodied AI interface realized as a wrist wearable made of machine knitting. Powered by LLM-driven affective inference, it is designed to bridge isolated people to social connection through tangible physical interface. Built at the MIT Physical AI Hackathon, the project makes the case that meaningful intelligence is not only computational, but physical: shaped by sensation, woven into material, expressed through tactiles.
The wearable captures continuous physiological signals including temperature, position, and pressure, and feeds them into an LLM-based affective inference pipeline that classifies emotional states in real time. An Arduino Uno Q integrates touch sensing, IMU, motor, and haptic actuator subsystems, closing the loop between body signal and embodied feedback.
Sensing
TemperaturePosition (IMU)PressureTouch
Hardware
Arduino Uno QPressure SensorDC MotorHaptic Actuator
AI Pipeline
LLM Affective InferenceMultimodal Input FusionEmotional State Classification
Materials
Knitted YarnPLA
Award
MIT Physical AI HackathonWinner · 2026
System
Sensing → Arduino Uno Q → LLM inference → haptic feedback
Documentation
