Predicting distress before crisis with Apple Watch HR and HRV

Physiological stress is visible in the data long before it's visible in behavior. Heart rate rises; heart-rate variability (HRV) — the tiny fluctuations between beats — collapses. The EchoSense Companion reads both from Apple Watch and turns them into something actionable.

Your baseline, not a population average

A resting heart rate of 85 is elevated for one person and normal for another. The app builds a personal baseline from each user's own history and measures deviation from that, so the score reflects this person on this day — not a textbook average.

Four tiers, bounded by design

• Calm — baseline range; the app stays out of the way.
• Elevated — early drift; Echo might gently surface a breathing exercise or a favorite story.
• Stressed — clear physiological load; regulation tools move front and center, and linked caregivers see the tier change.
• Crisis — sustained extreme signals; the app makes SOS one tap away and caregivers are alerted.

The score is deliberately bounded and interpretable — a tier, not an opaque 0-to-100 number. Caregivers and users can both understand why the app thinks what it thinks, and the tier can trigger gentle interventions instead of alarms.

The pipeline

Apple Watch streams HR and HRV through HealthKit; a Swift EventChannel feeds the Flutter app in near-real time; scoring happens on the phone. Vitals sync to a linked caregiver happens over an encrypted channel — but the raw analysis never needs a server.