Battery life in a medical wearable: a worked power budget
Battery life is the first question every client asks and the easiest one to get wrong. A simple duty-cycle power budget gives an honest answer early, before the enclosure is fixed.
Every component in a wearable draws current in a few states: on, idle and off. Battery life is just the battery capacity divided by the average current. The work is in finding that average honestly.
Step 1: list the states
Our reference patch, Hridya version 0.2, works in a 15-second cycle:
| State | Time | What is on | Current |
|---|---|---|---|
| Recording | 10 s | ECG, optical pulse and SpO₂, motion, Bluetooth streaming | 7–8 mA |
| Sensors off | 5 s | Processor idle, Bluetooth connected, no streaming | 3–5 mA |
Step 2: average over the cycle
The average current is each state’s current multiplied by its share of time:
- Conservative: (8 mA × 10 s + 5 mA × 5 s) ÷ 15 s = 7.0 mA
- Optimistic: (7 mA × 10 s + 3 mA × 5 s) ÷ 15 s = 5.7 mA
Step 3: divide the battery by it
With a 500 mAh cell, 500 ÷ 7.0 gives about 71 hours and 500 ÷ 5.7 about 88 hours: roughly three days. In practice, plan for 10 to 20 percent less, because cells lose capacity in the cold, with age and near the end of discharge.
Where the next gains are
The numbers above show where to work next. A 3 to 5 mA “sleep” is high for a Bluetooth processor that can idle in the microamp range. That current comes from parts that stay powered and from the radio staying in a fast connection. The usual steps are:
- Deeper sleep. Switch off sensor supplies fully and let the processor reach its low-power state between events.
- Longer connection intervals. Bluetooth LE can stay connected with packets every few hundred milliseconds instead of every few milliseconds when nothing urgent is sent.
- Send less. Buffer and batch samples, or process on the patch and send results instead of raw data, when the use allows it.
- Duty-cycle the light. The optical sensor’s LEDs are often the largest single load. Lower LED current and shorter pulses help, within signal-quality limits.
Measure, then publish
An estimate is a starting point. A real power budget comes from a current profiler on the bench and a full discharge test at body temperature. We label our own figures as calculated until that test is done, and we recommend clients ask any supplier the same question: measured or estimated?
Quick formula: battery life in hours = capacity in mAh ÷ average current in mA, then take off 10 to 20 percent for real-world losses.
Want an ECG and SpO₂ patch under your own brand?
Our prototype, Hridya, already records ECG, SpO₂ and motion and streams them to its own app. Develop it with us into your product, or let the same team build your own device.