U of A researchers use AI and wearable devices to predict labor through body temperature

TUCSON, Ariz. (KGUN) — A smartwatch may do more than track your steps: it might soon help predict when labor will start.

Researchers at the University of Arizona are using wearable sensors and artificial intelligence to detect the onset of labor, a development they say could transform pregnancy care.

Elise Erickson, a certified nurse-midwife and associate professor at the U of A College of Medicine, said the uncertainty around labor timing causes major challenges for families. Inspired by animal studies linking body temperature to birth, Erickson’s team tracked minute-by-minute body temperature using wearable rings.

Shravan Aras, who leads sensor analysis at the university, developed a deep-learning AI model that sifted through the massive data sets. The model was able to predict labor within a four-day window nearly a week in advance, with about 80% accuracy.

The team now hopes to integrate this technology into everyday wearables like Apple Watches and Fitbits, making predictive care more accessible.