Technical maintenance: some sensors may be temporarily unavailable. Support

Web3 public sensors map

Welcome to the decentralized opensource sensors map which operates with the sole intent of serving the free will of individuals , without any beneficiaries. It offers two distinct layers of decentralization at your choise: peer-to-peer connectivity for direct access to sensor data, and the federative concept for accumulating sensor data and displaying measurement history. Click here for further technical details.

Map data © OpenStreetMap contributors

airalab/sensors.social v4.1.0

Secured by

Your bookmarked sensors

Save any sensor from the map here for quicker access
← Back to Blog

Insight Sleep Analytics: Understanding Your Night Report and Comfort Score

Insight Sleep Analytics transforms bedroom environmental data (CO₂, temperature, humidity, PM2.5, and noise) into a nightly comfort report and score based on research-informed thresholds.


Your bedroom environment changes throughout the night — often in ways you don’t notice. Insight Sleep Analytics turns those subtle changes into a simple morning summary of how comfortable your sleep environment was.

The system evaluates:

  • CO₂ concentration
  • Temperature
  • Humidity
  • PM2.5 (when available)
  • Noise peaks (when available)

It then produces a Night Report and a Comfort Score that reflects overall environmental conditions during sleep.

The Sleep Score reflects environmental comfort, not sleep stages or medical sleep quality.


Target Ranges at a Glance

MetricGeneralBiohacking
CO₂≤ 750 ppm≤ 600 ppm
Temperature19–22°C17–20°C
Humidity40–60%40–50%
PM2.5≤ 5 µg/m³≤ 3 µg/m³
Noise≤ 5 hours with peak > 45 dB≤ 1 hour with peak > 45 dB

Your Night Report at a Glance

Altruist insight with night analytics screenAltruist insight with night analytics screen

The Sleep Analytics screen includes:

ElementMeaning
Large ring scoreOverall comfort score (general model)
Biohacking scoreStricter comfort model with tighter thresholds
Metric cardsNight summary per sensor (averages; noise = peak hours)

How Data Is Collected

Insight continuously reads environmental sensors:

  • Temperature
  • Humidity
  • CO₂

If Urban is enabled:

  • PM2.5 particles
  • Noise (noiseAvg / noiseMax)

Instead of storing raw high-frequency data, Insight computes hourly averages (and for noise, the hourly maximum of noiseMax), which are stored locally in device memory.

flowchart LR
    Sensors[Sensors] --> Hourly[Hourly averages / max]
    Hourly --> Storage[Local storage (~48h)]
    Storage --> Night[Night window]
    Night --> Score[Comfort Score]

No SD card is required — all required history is stored internally.


Defining the Night Window

The night period is configurable in the web interface:

  • Start (default): 22:00
  • End (default): 07:00 (exclusive)

This defines a typical sleep window of ~9 hours.

Special cases:

  • If start = end → 24/7 mode
  • If the night crosses midnight → data is automatically combined across two calendar days

When a Night Report is Generated

A report is only generated when enough data is available:

required_hours = ⌈2/3 × night_length⌉

For a 9-hour night:

  • Minimum required data: 6 hours

If insufficient data is available, the system shows:

Night data is collecting

along with the available / expected hours.


How Night Values Are Calculated

For most metrics:

  1. Sensor data is averaged per hour
  2. Only hours within the night window are selected
  3. A simple arithmetic mean is computed per metric

For noise, Insight counts peak hours: an hour is a peak if the maximum noiseMax sample in that hour is > 45 dB. The night card shows the number of such hours (not night-average dB).


The Comfort Score System

Insight converts environmental conditions into a single score from 0 to 100. There are two models:

1. Conservative Model

Designed for general users and broader comfort ranges.

2. Biohacking Model

Uses stricter thresholds for optimization-focused users.


Both models work in a similar way.

Your score starts at 100, which represents ideal sleeping conditions. Each environmental factor (such as CO₂, temperature, humidity, noise, and PM2.5) can slightly reduce this score when it moves away from its comfort range.

All of these small effects are combined into a single total “comfort impact”, which is then applied to the final score.

In simple terms:

Score = 100 − total discomfort (scaled)

score = clamp(100 + 2 × Σ(impact), 0, 100)

How Each Metric Affects the Score

Each environmental factor contributes independently to the final score.

CO₂ (Air Quality)

Elevated CO₂ levels are associated with reduced sleep quality and increased nighttime arousals.

ModelThresholdImpact
Conservative750 ppm−0.52% per +100 ppm
Biohacking600 ppm−0.80% per +100 ppm

Research basis: Controlled indoor studies show reduced deep sleep at elevated CO₂ levels (~2000 ppm conditions).


PM2.5 (Air Particles)

Fine particles are linked to inflammation and reduced sleep efficiency.

ModelThresholdImpact
Conservative5 µg/m³−0.3% per +10 µg/m³
Biohacking3 µg/m³−0.5% per +10 µg/m³

Research basis: WHO Air Quality Guidelines (2021), respiratory inflammation studies.


Noise (Peak Hours)

Night noise contributes to micro-arousals even when not consciously perceived. Insight scores how many hours had a peak above 45 dB, not the night-average level.

ModelAllowance (peak hours)Impact per excess peak
Conservative5−2.0%
Biohacking1−3.0%
excess = max(0, peak_hours − allowance)
impact = −pct_per_peak × excess

Research basis: WHO Environmental Noise Guidelines and sleep arousal literature (peaks matter more than a smoothed average).


Temperature

Thermal conditions strongly influence sleep onset and continuity. Both an upper and a lower bound apply.

ModelComfort bandImpact
Conservative19–22°C−1.5% per °C outside the band
Biohacking17–20°C−1.5% per °C outside the band

Research basis: Sleep thermoregulation studies (Walker, WHOOP datasets, PMC research).


Humidity

Humidity affects airway comfort and perceived air quality.

ModelRangeImpact
Conservative40–60%−0.2% per 10% outside range
Biohacking40–50%−0.4% per 10% outside range

Research basis: EPA indoor air quality recommendations and respiratory comfort literature.


Example

If your average CO₂ is 900 ppm:

  • Deviation above threshold (conservative): +150 ppm
  • Impact ≈ −0.52 × 1.5 = −0.78%

This is then combined with other metrics to form the final score.

A score of 100 means all metrics stayed within ideal comfort ranges (including noise peak allowance).


How Scores Are Interpreted

  • 90–100: Excellent environmental conditions
  • 70–90: Good, minor deviations
  • 50–70: Noticeable environmental stress
  • <50: Significant comfort issues

The score is designed as a feedback signal, not a medical metric.


Configuration

Altruist insight web-interface with night analytics configuration

In the web interface you can configure:

  • Night start/end times
  • Urban sensor integration
  • disable opening of sleep analytics screen each morning

Important Limitations

  • The device does not measure sleep stages or brain activity
  • Short spikes may not significantly affect hourly averages (temperature, humidity, CO₂, PM2.5)
  • Noise uses hourly max samples; very brief events can still register as a peak hour if noiseMax exceeds 45 dB
  • First-night data may be incomplete due to buffer initialization
  • The score is a sleep environment comfort signal, not a clinical assessment

Summary

Insight Sleep Analytics converts environmental conditions into a unified comfort score using research-informed thresholds and a weighted impact model.

It is still evolving. The model, thresholds, and experience will continue to be refined as we learn from real-world use.

If you’re using Insight, your feedback helps shape what comes next.

Try the device: Insight