A radar, not a camera
From the nightstand, a low-power radar notices presence, movement and the motion of breathing—through bedding, in complete darkness, with nothing worn. How the radar works ↓
Contactless bedside sensing
A room that knows when to help. And when to stay quiet.
From the nightstand, LunaWake reads your night—no camera, nothing to wear, nothing touching you. And its reading is being tested against polysomnography, the reference hospital sleep labs trust.
How it works
Sense. Understand. Respond—or stay quiet.
A signal is not a sleep state. LunaWake senses first, understands second, and only then responds—so a noisy signal never turns into a made-up sleep state.
The night, thirty seconds at a time
LunaWake is built to estimate when you are asleep, awake, or somewhere in between—in the same 30-second epochs sleep labs score. When the signal is unclear, it says so instead of guessing.
One response at a time
Light, sound and a gentle wake, chosen for the moment—and often the best response is none at all. Your choice always wins.
01 · The study
Tested against the sleep lab’s reference.
Polysomnography—PSG—is how sleep medicine scores a night: brain waves, eye movement, heart rhythm, airflow and breathing effort, read epoch by epoch by trained technicians. It is the reference any serious sleep technology has to face.
Study design
Nearly a hundred lab nights, dozens of sleepers. Every one scored by technicians.
Healthy adults slept at a university sleep laboratory; sleepers with disturbed breathing joined at a partner hospital’s sleep-medicine center. Each night was recorded two ways at once—full PSG on the body, LunaWake on the nightstand—and scored by technicians under AASM rules, the scoring manual of the American Academy of Sleep Medicine. The study is ongoing, and the count keeps growing. Results are shared here as they are finalized.
- 01Full head-electrode setup
- 02Eye-area placement detail
- 03Chest lead placement
- 04Leg electrodes and lead wires
The signal, up close
One night, side by side with the lab.
Statistics summarize; a night shows. Three windows from an early two-night feasibility session with one sleeper, recorded before the larger study above—LunaWake’s bedside reading against the lab’s own instruments. These are the breathing and movement signals the estimate reads from; they show the raw material, not a result.
Signal window 01 · Falling asleepTwo records, one minute—asleep at 23:37 by the lab’s scoring, and the bedside reading shifted at the same minute.See the chart
Respiration and movement are shown for 30 minutes on either side of the PSG-scored event.
Fell asleep
At 23:37, by the lab’s scoring—and the bedside reading shifted at the same minute. As this sleeper drifted off, breathing settled into a calmer, more even rhythm, and the two lines pulled to within one breath per minute.
Average difference (MAE) 2.5 → 0.8 breaths/min (≈3× closer) across PSG-scored sleep onset · 1 participant.
Signal window 02 · Overnight
Breathing rate tracks the lab’s chest belt across the displayed night.
One full night—LunaWake’s contactless reading, side by side with the PSG chest belt.
91%
of the night, the bedside reading stayed within three breaths a minute of the belt the lab straps to the sleeper’s chest—usually within a single breath. Close enough, on this night, to follow the sleeper’s breathing rhythm—with nothing touching them.
1 participant · night A shown · exploratory comparison
For the technically minded: average difference (MAE) 1.1 breaths/min · average bias +0.4 · ≈ 6 hours of compared sleep · single occupant, bedside range · pre-production algorithm.
Signal window 03 · Waking upThe clearest change of the night—awake at 06:59 by the lab’s scoring, and the bedside signal changed at the same minute.See the chart
Movement is shown through the PSG-scored awakening and out-of-bed event.
Woke up
At 06:59, by the lab’s scoring. On this night, waking ended the stillness of sleep—and the bedside signal changed at the same minute.
Movement index 2.3 → 17.4 (0–100 scale, 7.6×) after the PSG-scored final wake · 1 participant.
Heritage
The sleep-estimation approach builds on 500+ PSG-annotated nights from our partner laboratory’s prior sleep-sensing research.
Field data
Adapted to contactless radar with hundreds of LunaWake’s own overnight recordings from real bedrooms.
The test
Then held to nearly a hundred PSG-scored lab nights—healthy sleepers and clinic patients alike, with more still being recorded.
02 · The sensor
Radio waves, read at the millimeter.
Breathing moves the body by millimeters. A 60 GHz radar can measure that from the nightstand—here is how it works, and what it can and cannot see.
Reflections, not images
The radar sends out a faint radio signal and reads its reflection off the surface of the body. Breathing and movement shift that reflection by millimeters, and the radar measures the shift—through bedding, in complete darkness.
Bedside only
Its reach ends at the bedside: within 2.4 meters, about eight feet. There is no lens and no image. It cannot see a face or a room; it can only sense motion.
No camera, by design
The radar produces motion readings, not pictures. There is nothing to watch and nothing to leak from a lens, because there is no lens. Its microphones are used only for passive acoustic sensing, not voice commands or conversation, and the physical privacy switch cuts off acoustic sensing at the hardware level. What LunaWake records, and how it is stored, is described in our Privacy Policy.
03 · Response
A room that helps you settle—then gets out of the way.
Warm light and sound help you settle. As the night deepens, the room goes quiet, then brings you back gently when morning comes. You stay in control.
-
Settle in
Warm light. Slower sound. One guide, if you want it.
-
Drift off
As you settle, the room fades with you.
-
Sleep undisturbed
No prompts. No content. Just quiet sensing.
-
Wake gently
Light first. Sound only if needed.
-
Understand your morning
One useful insight—not another dashboard.
Change it, skip it, or turn it off anytime.
Evidence base
The ingredients draw from established research.
The evidence informs how the room uses light, sound and routine; it does not make an automated bedroom equivalent to clinical care.
A lower-energy environment
Guidance is informed by circadian light, paced-breathing and music research.
Chang 2015 · Brown 2022 · CIE S 026:2018 · Laborde 2022 · Jespersen 2022
Consistent cues, adjusted over time
Selected principles from CBT-I—cognitive behavioral therapy for insomnia—inform how the room supports your rhythm: a consistent wake anchor, clearer bed-and-sleep cues and gradual adjustments. LunaWake does not deliver CBT-I and is not a treatment for insomnia.
Qaseem 2016 · Edinger 2021 · Riemann 2023 · Espie 2012
Gradual light, then sound
Dawn-simulation and sleep-inertia research inform the sequence.
Trotti 2017 · Gabel 2013 · Thompson 2014 · McFarlane 2020
Where it stands
What’s tested. What’s held back.
LunaWake tells you what it knows and how it knows it—and when it doesn’t know, it says so. That rule shaped the study, and it shapes every night at home.
Tested
Sensing and sleep estimates evaluated against technician-scored PSG across nearly a hundred lab nights—healthy adults at a university sleep laboratory and patients at a hospital sleep-medicine center. Results are shared here as they are finalized.
Held back
When the signal is weak, your record shows a gap instead of a guess—and the room stays quiet rather than act on a weak read. Sleep estimates are wellness insights, not diagnosis.
Next
More sleepers, more homes, more positions. Shared here as results come.
A calmer bedroom, built on evidence
Be first to bring LunaWake into your night.
Join for product updates, early access and launch pricing when LunaWake becomes available.
Get early access and launch pricing ↗Science & Evidence
LunaWake is a general wellness product intended to support relaxation, sleep-related routines and self-awareness. It is not a medical device and is not intended to diagnose, treat, cure or prevent any disease or condition. Sleep estimates are wellness information, not medical measurements, and individual results vary. The laboratory study described on this page is ongoing algorithm research and development conducted with a university sleep laboratory and a hospital sleep-medicine center; it evaluates LunaWake’s sensing and algorithms against technician-scored polysomnography and is not a clinical trial of health outcomes. Third-party research cited on this page describes those studies, not LunaWake. Sensing is contactless and camera-free; what LunaWake records, and how it is stored, is described in our Privacy Policy. If you suspect a sleep disorder, consult a qualified professional.