← All 50 books
Book 36 of 50 · Free
Indoor Air Quality Monitoring with IoT
2,837 words · 17 chapters · illustrated
Indoor Air Quality Monitoring with IoT
Book 36 of 50 — AstolixGen Learning Series For researcher and publication students
About This Book
We spend 90% of our time indoors, breathing air that can be 2–5× more polluted than outside — yet it's invisible and unmeasured. This book builds complete indoor air-quality (IAQ) monitoring systems: the pollutants that matter, low-cost sensors and their calibration against reference instruments, ventilation control, and the health-evidence framing that makes IAQ papers publishable. You'll learn to produce trustworthy measurements, not just colorful dashboards.
Learning objectives: - Name the key indoor pollutants and their health-relevant thresholds (WHO 2021) - Select and calibrate low-cost gas/particulate sensors - Design multi-room monitoring deployments - Implement ventilation control from IAQ data - Handle sensor drift, cross-sensitivity, and humidity effects - Evaluate IAQ interventions with proper study design - Publish credible IAQ research
Chapter 1: Why Indoor Air Quality Matters
Indoor air carries particulate matter (PM2.5/PM10), CO₂, volatile organic compounds (VOCs), carbon monoxide, radon, and biological aerosols. Sources: cooking (the dominant PM source in homes), cleaning products, furniture off-gassing, candles, outdoor infiltration, and occupants themselves (CO₂, bioeffluents). Health effects range from headaches and lost productivity (CO₂ > 1000 ppm impairs cognition measurably) to chronic cardiovascular and respiratory disease from long-term PM2.5 exposure.
The WHO's 2021 global air quality guidelines tightened PM2.5 limits (annual mean 5 µg/m³) — levels many indoor environments exceed during cooking. For researchers, IAQ is attractive: the problem is real, the measurements are physical, interventions (ventilation, purifiers) are testable, and the literature spans environmental science, building engineering, and public health — a genuinely interdisciplinary publication space.
Example: A Karachi classroom study: PM2.5 spiked to 180 µg/m³ during nearby cooking hours (infiltration), CO₂ exceeded 2000 ppm by midday in a 40-student room. Two interventions tested: scheduled window ventilation (CO₂ −45%) and a DIY box-fan purifier (PM2.5 −70%). Both measured, both published.
For your research: Frame IAQ papers around a health-relevant threshold (WHO guideline), not just "we measured pollution." Threshold-crossing hours per day is a powerful, interpretable metric.
Key takeaway: Indoor air is often worse than outdoor; anchor studies to WHO 2021 thresholds and threshold-exceedance metrics.
Chapter 2: The Pollutants — What to Measure
PM2.5/PM10 (particulate matter): the most health-relevant. Low-cost optical sensors (Plantower PMS series, Sensirion SPS30) count particles via laser scattering. CO₂: occupancy and ventilation proxy; NDIR sensors (Sensirion SCD30/SCD40, MH-Z19) are accurate and stable. VOCs: from cooking, cleaning, materials; metal-oxide (MOX) sensors (SGP30/SGP40, BME680) respond to many gases but are non-specific — report as "VOC index," never as specific compounds. CO: safety-critical near combustion; electrochemical sensors. Temperature/humidity: essential context — PM sensors are humidity-sensitive, and comfort analysis needs both.
What NOT to claim: low-cost MOX sensors cannot identify formaldehyde specifically; optical PM sensors misread humidity as particles. Papers that overclaim sensor specificity get rejected. Report what the sensor actually measures.
Example sensor set per room (~$60): PMS5003 (PM), SCD40 (CO₂/temp/RH), SGP40 (VOC index). Calibrated against a reference (see Ch. 3). This trio covers 90% of IAQ research questions.
For your research: Your methods must state each sensor's principle, range, and known cross-sensitivities. A sensor table is expected.
Key takeaway: PM + CO₂ + VOC-index + temp/RH covers most IAQ work; never overclaim specificity.
Chapter 3: Calibration — Making Low-Cost Data Trustworthy
Raw low-cost sensor data is not trustworthy. PM sensors overread in high humidity (water droplets scatter light like particles) — apply humidity correction (published formulas exist) or dry the inlet. CO₂ NDIR sensors are stable but need periodic auto-calibration (ABC assumes weekly fresh-air exposure; disable ABC in always-occupied rooms and calibrate manually). MOX VOC sensors drift with age and respond to humidity — use relative indices, not absolute ppb.
Calibration procedure: co-locate your low-cost node with a reference instrument (borrowed from an environmental lab, or a calibrated mid-range device) for 1–2 weeks across varying conditions. Fit a correction (linear or multilinear with humidity/temperature terms). Report R², RMSE, and bias before/after. Re-check quarterly — publish the drift.
Example: PMS5003 vs reference (TSI DustTrak): raw R² 0.82, RMSE 14 µg/m³; after humidity correction: R² 0.94, RMSE 6 µg/m³. The correction equation and validation stats were a full paper section.
For your research: Co-location calibration with reported metrics is the minimum bar for credible low-cost IAQ papers. No calibration section = desk rejection in serious venues.
Key takeaway: Co-locate, correct (especially humidity), report R²/RMSE, re-check drift — calibration is the paper's foundation.
Chapter 4: Deployment Design — Rooms, Placement, Density
Placement: breathing zone height (1–1.5 m), away from direct sources (not above the stove), away from windows/doors (drafts), one node per room for studies (hallway nodes miss room-level variation). Density: for a home study, 3–5 nodes (bedroom, living, kitchen, outdoor reference). The outdoor node is essential — it separates indoor sources from infiltration.
Power/connectivity: mains-powered nodes (sensors like PMS draw too much for batteries); Wi-Fi or LoRa to a gateway; local buffering for outages (SD card). Duration: minimum 2 weeks per condition; seasonal variation means winter and summer campaigns differ enormously.
Example: A 20-home study: 4 indoor nodes + 1 outdoor per home, 3-week winter campaign. Key finding: kitchen PM during cooking exceeded outdoor pollution by 8×, but 70% of daily exposure came from the living room (time-weighted) — placement + occupancy diaries changed the conclusion.
For your research: Include occupancy/activity diaries (even simple ones). IAQ without knowing what people were doing is hard to interpret — and reviewers know it.
Key takeaway: Breathing-zone placement, outdoor reference node, activity diaries, multi-week campaigns.
Chapter 5: CO₂ and Ventilation — The Actionable Loop
CO₂ is the most actionable IAQ metric: it directly indicates ventilation adequacy, and the fix (more fresh air) is straightforward. Thresholds: <800 ppm good, 800–1200 acceptable, >1200 poor (cognitive effects documented), >2000 seriously under-ventilated. Demand-controlled ventilation: CO₂ sensor → controller → damper/fan/window alert. In classrooms, this single loop is the highest-impact IAQ intervention.
Example: A 35-student classroom: baseline CO₂ peaked at 2,400 ppm. Intervention: CO₂-triggered exhaust fan at 1,000 ppm + teacher alert. Result: peaks capped at 1,100 ppm, absenteeism (illness) dropped 12% over the term vs the control classroom. The control classroom is what made it a paper, not a project.
For your research: Ventilation studies need a control room and should report both CO₂ and an outcome (absenteeism, test scores, or at least perceived air quality surveys).
Key takeaway: CO₂ → ventilation is the highest-ROI IAQ loop; evaluate with controls and outcomes.
Chapter 6: Particulates — Sources, Purifiers, and Evidence
Indoor PM sources ranked: cooking (frying = extreme), candles/incense, cleaning (resuspension), outdoor infiltration (traffic, burning), smoking. Interventions: source control (lids on pans, range hoods that actually vent outside), portable HEPA purifiers (CADR matched to room size), and DIY box-fan + MERV-13 filters (the famous Corsi-Rosenthal box — ~$60, remarkably effective, great for low-resource studies).
Measuring purifier effectiveness: decay-rate method — elevate PM (or wait for a cooking event), run purifier, fit exponential decay; compare against natural decay (no purifier). Report clean air delivery in context of room volume.
Example: Corsi-Rosenthal box in a 30 m³ bedroom: PM2.5 decay constant improved 3.2× vs natural decay; overnight mean PM2.5 fell from 34 to 11 µg/m³. Total cost $58. The cost-effectiveness framing drove citations.
For your research: Low-cost interventions with rigorous measurement are highly publishable and highly cited — they serve the low-resource settings most papers ignore.
Key takeaway: Source control first, purifiers second; measure with decay rates; cost-effectiveness is a publication asset.
Chapter 7: Data Analysis for IAQ
IAQ analysis patterns: time-weighted exposure (concentration × time per microenvironment — the health-relevant dose), event detection (cooking peaks via PM slope), source apportionment (indoor/outdoor ratios by time of day), threshold exceedance (hours/day above WHO limits), and intervention comparison (before/after with controls, or crossover designs where each home serves as its own control).
Crossover design (powerful for small-N studies): each home gets 2 weeks baseline + 2 weeks intervention, order randomized. Each home is its own control — eliminating between-home variation. With 10 homes you have a real study.
Example: Purifier crossover in 12 homes: baseline week mean PM2.5 28 µg/m³ → purifier week 12 µg/m³ (p < 0.001, paired test). The paired design gave statistical power that 12 homes in a parallel design couldn't.
For your research: Learn the crossover design — it's the small-N researcher's superpower for intervention studies.
Key takeaway: Time-weighted exposure, threshold exceedance, and crossover designs are the IAQ analyst's core tools.
Chapter 8: Sensor Networks and IoT Architecture for IAQ
Apply Books 31/34: nodes publish via MQTT (QoS 1, retained last values), a broker bridges to a time-series DB (InfluxDB), Grafana dashboards per home, alerts on threshold exceedance. IAQ-specific needs: high time resolution during events (1-min), aggressive downsampling for storage (1-min → 1-hour rollups after 30 days), and data-quality flags (sensor warm-up periods, humidity-suspect PM readings).
Privacy: indoor data reveals occupancy patterns — encrypt, restrict access, anonymize in publications (Home A, not addresses), get consent. Ethics approval is required for human-occupancy studies at most institutions.
Example architecture: ESP32 + sensors → MQTT → Mosquitto → Telegraf → InfluxDB → Grafana; 5 nodes/home × 20 homes; total cloud cost $0 (local server). The zero-cloud-cost design was a deliberate choice for a low-resource setting, stated in the paper.
For your research: State the full stack with versions; include the ethics approval ID; anonymization procedures belong in methods.
Key takeaway: MQTT + time-series DB + Grafana is the standard IAQ stack; ethics and anonymization are mandatory.
Chapter 9: Health Evidence — Linking Air to Outcomes
Strong IAQ papers connect exposure to outcomes: cognitive (test scores, reaction time vs CO₂), respiratory (symptom diaries, peak-flow vs PM), absenteeism, sleep quality (bedroom PM/CO₂ vs sleep trackers). You don't need clinical trials — validated symptom questionnaires and simple cognitive tests suffice for a first paper.
Caution: correlation is not causation; control for confounders (season, illness circulation, exam stress). Report effect sizes with confidence intervals, not just p-values.
Example: Classroom CO₂ vs math test scores (n=180 students, 6 classrooms): each 500 ppm increase associated with −3.2% scores (95% CI −5.1 to −1.3), controlling for temperature and time of day. Modest, honest, publishable.
For your research: Partner with a health or education researcher — interdisciplinary co-authors strengthen both the study and its reception.
Key takeaway: Link exposure to measured outcomes with confounder control; interdisciplinary co-authors help.
Chapter 10: Low-Resource Settings — The High-Impact Frontier
Most IAQ research comes from wealthy countries; the worst exposures are elsewhere (biomass cooking, brick kilns, traffic). Low-cost sensors + DIY purifiers + local deployments = high-impact research. Challenges: power reliability (design for outages), dust (sensor maintenance schedules), and community engagement (results must return to participants in understandable form).
Example: A Lahore kitchen study: biomass stove PM2.5 averaged 210 µg/m³ during cooking; a $12 chimney hood intervention cut it to 68 µg/m³. The paper's community-reporting section (results shared in Urdu infographics) was praised by reviewers and adopted by an NGO.
For your research: Low-resource IAQ work is under-published and over-cited. Budget for community reporting — it's both ethical and strategically smart.
Key takeaway: The highest-impact IAQ research serves low-resource settings; report back to communities.
Chapter 11: Standards, Guidelines, and Policy Context
Know the landscape: WHO 2021 global air quality guidelines (PM2.5 annual 5 µg/m³, 24-h 15 µg/m³), ASHRAE 62.1 (ventilation rates), WELL/RESET building standards (commercial IAQ certification), and national building codes. Your paper's discussion should position findings against these: "bedroom CO₂ exceeded the 1,000 ppm guideline for 4.2 h/day" is policy-relevant language.
Example: A school-district study framed entirely around ASHRAE ventilation rates led to a district-wide ventilation retrofit policy — the paper's policy impact section documented it, multiplying the work's influence beyond citations.
For your research: One paragraph linking results to a specific standard or policy turns a measurement paper into a policy-relevant one.
Key takeaway: Anchor discussions to WHO/ASHRAE/WELL — policy relevance multiplies impact.
Chapter 12: Your IAQ Study — From Question to Paper
Template:
- Question: e.g., "Does a low-cost purifier reduce children's bedroom PM2.5 exposure in Lahore homes?"
- Design: Crossover (each home its own control), 2+2 weeks, n≥10 homes.
- Instruments: Listed sensors with calibration (co-location stats), placement map, outdoor reference.
- Metrics: Time-weighted PM2.5, threshold-exceedance hours, pre-registered.
- Analysis: Paired tests, effect sizes, confounder checks.
- Ethics: Approval ID, consent, anonymization.
- Economics: Intervention cost, cost per µg/m³ reduced.
- Artifact: Dataset (anonymized), calibration data, analysis code.
- Writing: IEEE or environmental-health journal format; abstract leads with the exposure reduction numbers.
Common rejections: no calibration, no control/crossover, N too small without paired design, overclaimed sensor specificity, missing ethics.
For your research: This is A1→A4 once more — and IAQ's physical measurements plus crossover designs make it one of the most achievable first-publication domains.
Key takeaway: Calibrated sensors + crossover design + threshold metrics + ethics + artifact = a publishable IAQ paper.
Learning Dashboard
| # | Chapter | Core idea | Research use |
|---|---|---|---|
| 1 | Why IAQ | 90% indoors, WHO 2021 | Threshold-exceedance metrics |
| 2 | Pollutants | PM/CO₂/VOC, sensor limits | Sensor table, no overclaiming |
| 3 | Calibration | Co-location, humidity correction | R²/RMSE reporting = credibility |
| 4 | Deployment | Placement, outdoor ref, diaries | Multi-week, activity context |
| 5 | CO₂/ventilation | Actionable loop | Control-room evaluations |
| 6 | Particulates | Sources, purifiers, decay rates | Cost-effectiveness framing |
| 7 | Analysis | Exposure, crossover designs | Small-N superpower |
| 8 | IoT architecture | MQTT→TSDB→Grafana | Full-stack disclosure, ethics |
| 9 | Health links | Outcomes with confounder control | Interdisciplinary co-authors |
| 10 | Low-resource | High impact, community reporting | Under-published, over-cited |
| 11 | Standards | WHO/ASHRAE/WELL anchoring | Policy relevance |
| 12 | Study template | Question→paper pipeline | A1–A4 mapping |
References
[1] World Health Organization, WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide, Geneva, 2021. [2] P. Kumar et al., "Real-time sensors for indoor air monitoring and challenges ahead in deploying them to urban buildings," Science of the Total Environment, vol. 560–561, pp. 150–159, 2016. [3] L. Morawska et al., "Applications of low-cost sensing technologies for air quality monitoring and exposure assessment: How far have they gone?" Environment International, vol. 116, pp. 286–299, 2018. [4] J. Li, S. K. R. Haugen, and M. C. Turner, "Low-cost air quality sensors: Intercomparison and calibration," Atmospheric Environment, 2019. (verify exact citation) [5] U. Satish et al., "Is CO₂ an indoor pollutant? Direct effects of low-to-moderate CO₂ concentrations on human decision-making performance," Environmental Health Perspectives, vol. 120, no. 12, 2012. [6] R. Allen, P. Wargocki et al., "Associations of cognitive function scores with carbon dioxide, ventilation, and volatile organic compound exposures in office workers," Environmental Health Perspectives, vol. 124, no. 6, 2016. [7] A. Banks and R. Gupta, "MQTT Version 3.1.1," OASIS Standard, Oct. 2014. [8] L. Atzori, A. Iera, and G. Morabito, "The Internet of Things: A survey," Computer Networks, vol. 54, no. 15, pp. 2787–2805, 2010. [9] W. J. Fisk, "The ventilation problem in schools: literature review," Indoor Air, vol. 27, no. 6, 2017. [10] ASHRAE, ANSI/ASHRAE Standard 62.1-2022, Ventilation for Acceptable Indoor Air Quality, 2022.
Glossary
- PM2.5/PM10 — particles ≤2.5/10 µm; key health metric
- NDIR — nondispersive infrared; CO₂ sensing principle
- MOX — metal oxide; broad-spectrum gas sensing
- VOC — volatile organic compounds
- Co-location — calibrating against a reference side by side
- Crossover design — each subject gets both conditions in random order
- Time-weighted exposure — concentration × time, the health-relevant dose
- CADR — clean air delivery rate (purifiers)
- Corsi-Rosenthal box — DIY box-fan + filter purifier
- ABC — automatic baseline calibration (CO₂ sensors)
- Threshold exceedance — hours above a guideline limit
Practice Exercises
- Why is threshold-exceedance (hours/day above WHO limits) a better metric than daily mean? Give an example.
- Your PM sensor reads high on humid days. Explain the physics and two corrections.
- Design a 4-node home deployment: rooms, placement heights, and why each location.
- Write the control logic for CO₂-triggered ventilation with hysteresis (to avoid flapping).
- Describe the decay-rate method for purifier testing. What is the control condition?
- Design a crossover study for a purifier intervention with 10 homes. State the analysis.
- Why must MOX VOC sensors be reported as an index, not specific compounds?
- Compute time-weighted PM2.5 exposure for a person: 8 h bedroom at 12 µg/m³, 8 h office at 35 µg/m³, 2 h kitchen at 90 µg/m³, 6 h outdoors at 45 µg/m³.
- List the ethics requirements for an occupied-home IAQ study at your institution.
- Draft an abstract for your IAQ study with the key numbers filled in as variables.
End of Book 36. Next: Book 37 — IoT Security Basics.