Mammal smart garment monitors breastfeeding with 5% error on latch duration
A wearable bra detects infant heart rate and feeding patterns without any baby sensors.
Mammal is a caregiver-worn computational garment designed to unobtrusively monitor breastfeeding without any sensors attached to the infant. The system leverages inter-body signal transmission: natural mouth-to-breast contact creates a conductive path that allows the garment to capture the infant's cardiac signals and feeding-related acoustic vibrations on the caregiver's body. Custom algorithms then detect latch onset, infer infant electrocardiogram (ECG), and distinguish suck and swallow events. This enables estimation of latch duration, in-feeding heart rate, suck-swallow-breathe (SSB) ratio, and milk intake -- all from a comfortable, wearable form factor.
In a user study with 10 caregiver-infant dyads, Mammal demonstrated strong performance metrics: a mean absolute percentage error (MAPE) of 5.56% for latch duration, mean absolute error (MAE) of 3.61 bpm for infant heart rate, MAE of 0.12 for SSB ratio estimation, and a mean relative error of 15.76% for milk intake. Participants reported high comfort and wearability, suggesting the system could be integrated into daily life. Submitted to IMWUT'26, this research provides a practical solution for objective breastfeeding monitoring, potentially aiding lactation consultants, pediatricians, and parents in tracking infant feeding competence and health.
- Mammal is caregiver-worn and requires no sensors on the baby; data is captured via inter-body signal transmission through mouth-to-breast contact.
- Achieves 5.56% error on latch duration, 3.61 bpm heart rate accuracy, and 15.76% relative error on milk intake in a 10-dyad study.
- Detects latch onset, infant ECG, suck/swallow events, and estimates SSB ratio for comprehensive feeding monitoring.
Why It Matters
Enables objective, non-invasive breastfeeding monitoring for better infant health insights and parental peace of mind.