Insight Sleep Analytics: Understanding Your Night Report and Comfort Score

---

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 level (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.

---

Your Night Report at a Glance

The Sleep Analytics screen includes:

Element	Meaning
Large ring score	Overall comfort score (general model)
Biohacking score	Stricter comfort model with tighter thresholds
Metric cards	Night averages for each sensor

---

How Data Is Collected

Insight continuously reads environmental sensors:

• Temperature
• Humidity
• CO₂

If Urban is enabled:

• PM2.5 particles
• Noise level

Instead of storing raw high-frequency data, Insight computes hourly averages, which are stored locally in device memory.

flowchart LR
    Sensors[Sensors] --> Hourly[Hourly averages]
    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 Averages are Calculated

Insight does not evaluate raw sensor spikes. Instead:

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

This ensures stable and robust nightly values.

---

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.

Model	Threshold	Impact
Conservative	750 ppm	−0.52% per +100 ppm
Biohacking	600 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.

Model	Threshold	Impact
Conservative	5 µg/m³	−0.3% per +10 µg/m³
Biohacking	3 µg/m³	−0.5% per +10 µg/m³

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

---

Noise

Night noise contributes to micro-arousals even when not consciously perceived.

Model	Threshold	Impact
Conservative	35 dB	−2.5% per +10 dB
Biohacking	30 dB	−3.5% per +10 dB

Research basis: WHO Environmental Noise Guidelines and sleep arousal literature.

---

Temperature

Thermal conditions strongly influence sleep onset and continuity.

Model	Threshold	Impact
Both	+25°C / +20°C	−1.5% per +1°C above threshold

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

---

Humidity

Humidity affects airway comfort and perceived air quality.

Model	Range	Impact
Conservative	40–60%	−0.2% per 10% outside range
Biohacking	40–50%	−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.

---

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

In the web interface you can configure:

• Night start/end times
• Urban sensor integration

---

Important Limitations

• The device does not measure sleep stages or brain activity
• Short spikes may not significantly affect hourly averages
• 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