Wearable electronic skin tracks nine physical and sweat signals to classify stress responses
What was demonstrated
A Caltech-led team built a wrist-worn electronic skin that measures three vital signs (pulse waveform, galvanic skin response and skin temperature) and six sweat biomarkers (glucose, lactate, uric acid, sodium, potassium and ammonium). Its electrochemical sensors gave stable sweat analysis for over 100 hours, and it recorded continuously over 24 hours during daily activities in human participants. A machine learning pipeline distinguished three stressors with 98.0% accuracy.
Why it matters
Combining physical and chemical signals in one patch gives a more objective way to assess stress responses than questionnaires, and the sensor-stabilisation approach could transfer to other wearable biosensors.
Caveats
- Stressor classification was tested on small, controlled study groups; real-world clinical validation is still needed.
- Sweat biomarker readings depend on sweat rate and may not track blood levels closely.
- Accuracy figures come from the study's own data set and machine learning pipeline.
Technology profiles
Sources
- peer reviewed paper
A physicochemical-sensing electronic skin for stress response monitoring
Changhao Xu, Yu Song, Juliane R. Sempionatto et al.. Nature Electronics, 19 Jan 2024. DOI: 10.1038/s41928-023-01116-6.
Last reviewed 20 Oct 2026.