Smart sleep collar

Sleep loss costs the UK economy around £40 billion a year. Scalable, comfortable, and durable solutions are urgently needed to measure sleep in the real world. A potential game-changer in wearable sleep tech has emerged from the Luigi G. Occhipinti lab at the Division of Electrical Engineering, University of Cambridge which combines a smart 'collar' with deep learning neural networks and explainable AI. This versatile solution can classify 6 different sleep states with 98.6% accuracy using subtle vibrations from throat muscle activity. Not only does this hold great potential for a broad consumer market, such a tool could offer new insights into the fundamentals of sleep and enable extraction of more accurate sleep data than wrist-worn wearables, particularly for patients with abnormal or restricted limb movements. Interpreting sleep data from wearables can be a pain in the neck; this e-textile gives it a hug.

Check out summary here

And PNAS paper here