Skin tone, Confidence, and Data Quality of Heart Rate Sensing in WearOS Smartwatches
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Smartwatches can collect heart rate data unobtrusively and continuously, making them a promising tool for conducting long term studies, monitoring chronic conditions, and providing timely intervention. Healthcare applications, however, require us to understand the reliability of collected readings, both in terms of quality and quantity. The accuracy of optical heart rate (HR) measurements has been studied extensively in recent years, identifying several common causes of errors. For example, previous research has demonstrated that inaccurate HR readings occur more frequently in dark skin as compared to light skin due to melanin absorption. Smartwatches therefore implement a confidence mechanism to estimate reliability of HR readings. We study the effect of skin tone on the reliability of confidence estimation of seven consumer-grade WearOS smartwatches. We find that some watches systematically underestimate the reliability of HR readings taken from dark skin, despite no substantial difference in actual error. This results in significantly fewer data points for people with darker skin tones, which can bias downstream applications. We also report a wide variation in how watches implement the same WearOS API for HR collection, with implications for researchers that intend to use them for studies.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it