How AIVELA Detects Sleep and Estimates Sleep Stages
AIVELA does not know you are asleep simply because you stopped moving. It looks at patterns across time, combining movement with nighttime physiological signals to identify when sleep likely began, when you woke, and how the night may have moved through Wake, Light, Deep, and REM sleep. These estimates can help you understand sleep duration, continuity, timing, and structure. But a ring does not measure brain waves, eye movements, or muscle activity, so its sleep stages are not the same as a clinical sleep study.
Sleep is more than stillness
You may lie still while reading, watching something, or trying to fall asleep. You may also move while genuinely asleep. If a wearable used movement alone, quiet wakefulness could look like sleep and restless sleep could look like wakefulness.
AIVELA therefore considers several kinds of information together:
- movement and stillness from the motion sensor;
- pulse and pulse-timing patterns from optical PPG;
- changes in resting heart rate and other nighttime physiological signals;
- the timing, duration, and continuity of the detected period.
No single signal proves sleep. The algorithm looks for a pattern that becomes more convincing when the signals agree across time.
From a possible sleep period to your main sleep
The first task is to find periods that look like sleep rather than ordinary inactivity. The system then considers their timing, length, overlap, and gaps to form a sleep report.
Your main sleep is the sleep period used for the primary nightly result. It may be one continuous period or a combined view when qualifying sleep periods belong to the same larger sleep episode. Time awake between combined periods remains awake time; it is not silently converted into sleep.
Shorter sleep can still matter, but AIVELA handles it differently from your main sleep.
How AIVELA handles naps
AIVELA can detect naps lasting 20 minutes or longer. A qualifying nap contributes to Calm, giving the day a more complete view of rest and recovery.
For most naps, AIVELA shows the detected start and end time without dividing the nap into Light, Deep, or REM. In a short sleep period, detailed stage labels add little practical value and can suggest more certainty than the signals support.
Nap detection is still an algorithmic estimate. Prolonged sitting can sometimes resemble sleep—especially when your hand remains completely still, your body barely moves in the chair, and your heart rate stays steady. In that situation, quiet inactivity may occasionally be recognized as a nap.
How AIVELA estimates sleep stages
Sleep changes throughout the night. Movement, pulse rate, pulse timing, breathing-related patterns, and other peripheral signals tend to shift as the body moves between wakefulness and different stages of sleep.
The algorithm examines these patterns in successive time windows and assigns the stage that best matches the available evidence:
- Wake — periods that appear awake within the sleep interval;
- Light — lighter non-REM sleep;
- Deep — deeper slow-wave sleep;
- REM — rapid eye movement sleep.
The result is a hypnogram: a timeline showing how the estimated stages changed through the night. Real sleep is continuous and transitions are not always sharp, so two neighboring windows may be difficult to distinguish. A wearable has to choose the most likely label even when the signals sit near a boundary.
AIVELA is strongest at identifying the beginning and end of sleep, which are highly accurate. Sleep efficiency and the proportion of the night spent in Deep sleep are also relatively accurate. The exact stage labels—and the sequence of those stages across the night—can differ from PSG or another wearable. That is normal: PSG observes brain, eye, and muscle activity directly, while wearables infer sleep from peripheral signals, and different devices use different algorithms and decision boundaries.
What AIVELA calculates from the night
Once a main sleep period is available, AIVELA can organize the night into several useful measures:
- Time in bed covers the main sleep period from beginning to end.
- Total sleep time adds Light, Deep, and REM, excluding Wake.
- Wake after sleep onset adds the awake periods after sleep began.
- Sleep efficiency compares total sleep time with time in bed.
- Sleep timing considers when your main sleep began relative to your personal pattern.
- Sleep movement describes how much movement appeared during the main sleep.
Calm brings these measures together with Deep and REM estimates and nighttime resting heart rate. It answers a broader question than “Did I sleep?”: did the main sleep appear sufficient, continuous, well-timed, and supportive of recovery?
Why your result may not match your memory
Your experience and the ring are observing sleep from different places. You may not remember brief awakenings. You may remember being awake while lying still, which can be difficult for a wearable to distinguish from Light sleep. A late nap, split sleep, shift work, an unusual schedule, restless movement, poor contact, or interrupted data can also change the report.
A mismatch does not automatically mean your experience or the device is wrong. Start by asking which part differs: bedtime, wake time, total sleep, awake time, or a particular stage. Each difference has different possible explanations.
How to read sleep stages well
- Begin with the whole night. Sleep start and end times are among AIVELA's most accurate sleep results, so first check whether the main sleep period matches your night.
- Read duration and continuity before the stage timeline. Sleep efficiency and the overall proportion of Deep sleep are more dependable than the exact placement of every stage.
- Look for repeated patterns. One night can change with routine, environment, and measurement quality.
- Check fit and wear. Stable contact and consistent overnight wear give the algorithm a better signal.
- Use stages as estimates. A night with more or less Deep or REM than expected is not, by itself, evidence of a sleep disorder.
The goal is not to achieve a perfect stage chart. It is to understand how your sleep behaves over time and whether changes align with what you experience.
How wearable sleep differs from polysomnography
Clinical polysomnography, or PSG, measures brain electrical activity, eye movements, muscle activity, heart signals, breathing, and other information. Sleep specialists use those signals to score sleep stages and evaluate sleep disorders.
AIVELA RING PRO measures peripheral signals at the finger. Its sleep and stage results are algorithmic estimates for consumer wellness. Differences in exact stage classification and timing compared with PSG—or with another wearable—are expected and do not by themselves mean that one result is defective. AIVELA has not published evidence that its sleep-stage timeline is equivalent to PSG, and Oura’s validation results cannot be transferred to AIVELA.
A ring cannot diagnose insomnia, sleep apnea, parasomnias, or another sleep disorder. If you have persistent sleep difficulty, loud snoring or breathing concerns, unusual nighttime behavior, significant daytime sleepiness, or another medical concern, speak with a qualified healthcare professional.
Wearable sleep stages are a map of patterns around sleep—not a direct view inside the sleeping brain.