Prototype glasses use facial movement to generate sensing signals
Northwestern University engineers have developed prototype glasses that use facial movements to generate electrical sensing signals, allowing the system to distinguish activities including blinking, jaw clenching, eating and talking.
GlassTENG incorporates three thin triboelectric sensors around the frame. Contact and separation between material layers produce a voltage as the wearer moves.
Sensors at the nose bridge, temple and near the ear detect different combinations of arterial pulse and facial movement. In a study involving 20 participants, machine-learning analysis distinguished six activities with an overall accuracy of 93.8%. [Source 1]
Northwestern reports combined power consumption of 4.1µW for the three sensors. That figure does not establish the power budget of the complete system, which also streams data over Bluetooth.
The research was led by Nivedita Arora. On-device computation is a future development goal, rather than a demonstrated feature of the current prototype.
The work is included in the Proceedings of the 2026 ACM International Symposium on Wearable Computers and is scheduled for presentation in its Eyewear Computing session on 15 October. [Source 2]
