Sensor Framework Flags When Opening Windows Helps or Hurts Indoor Air

Opening a window can flush out indoor-generated pollutants, but the same move can pull outdoor fine particulate matter into a home. Researchers have now built a sensor-driven screening framework that decides, hour by hour, when natural ventilation looks like an opportunity rather than a risk, combining sustained outdoor PM2.5 measurements with a pragmatic temperature and humidity envelope.

The team used harmonised AirGradient observations accessed through OpenAQ to compare four residential communities: KL_TTDI in Kuala Lumpur, Malaysia; ID_BogorSelatan in Bogor, Indonesia; SG_Midwood in Singapore; and AU_Palmerston in Gray/Palmerston, Australia. The analysis covered 1 January to 1 February 2026 in local time, with 724 to 744 hourly bins per site. Sustained PM2.5 was represented by a trailing 24-hour mean and evaluated against a threshold of 15 micrograms per cubic metre, matching the 2021 World Health Organization 24-hour air-quality guideline value. The meteorological screen required temperatures between 24 and 30 degrees Celsius and relative humidity at or below 65 percent.

Only hours passing both tests counted as candidate natural-ventilation opportunities. The resulting clean-hour fractions differed sharply across sites: 14.5 percent in KL_TTDI, 34.3 percent in ID_BogorSelatan, 30.8 percent in SG_Midwood, and 97.4 percent in AU_Palmerston. Standalone thermal-humidity feasibility told a different story, at 97.0, 71.5, 98.5, and 20.6 percent respectively, revealing that the binding constraint varies by location.

Those paired numbers define each community’s monthly ventilation regime. Kuala Lumpur and Singapore were PM2.5-constrained: the weather was almost always comfortable enough to open windows, but outdoor particulate levels frequently blocked the opportunity. Bogor was mixed, with both pollution and climate each removing a substantial share of hours. Palmerston was the inverse case, predominantly clean by the PM2.5 measure but T-RH-limited, since its hot, humid conditions rarely fell inside the comfort envelope.

The authors are careful about scope. The framework identifies outdoor screening opportunities only; it does not measure what actually happens indoors. Because opening windows couples a dwelling to outdoor air, paired indoor-outdoor monitoring is needed before any translation into occupant guidance or automated control. The study is also a one-month, four-site comparison using low-cost sensor networks, so the fractions describe that specific January 2026 window rather than long-term climatological patterns.

Methodologically, the work is a Brief Report presenting a state-space screening approach rather than a clinical or intervention trial. Its value lies in showing how a simple, transparent rule set—trailing 24-hour PM2.5 against the WHO guideline value, plus a fixed temperature and humidity band—can turn raw sensor streams into interpretable regime classifications. The contrast between the 97.4 percent clean-hour fraction in Palmerston and the 14.5 percent in Kuala Lumpur illustrates how the same global guideline can produce radically different local ventilation calendars.

In T-RH-limited settings, the reverse holds. The researchers emphasise that these are screening outputs, not recommendations, and that validating them against measured indoor concentrations is the necessary next step before the approach could inform ventilation decisions in homes.

Subject of Research: A sensor-driven framework screening hourly natural-ventilation opportunities using outdoor PM2.5 and thermal-humidity constraints across four tropical and subtropical communities.

Article Title: A Sensor-Driven State-Space Framework for PM2.5-Informed Natural Ventilation Opportunities Under Coupled Pollution–Thermal–Humidity Constraints

Article References: Yu, Z., Deng, Y., & Yu, H. (2026). A Sensor-Driven State-Space Framework for PM2.5-Informed Natural Ventilation Opportunities Under Coupled Pollution–Thermal–Humidity Constraints. Aerosol and Air Quality Research. https://doi.org/10.1007/s44408-026-00167-6

Image Credits: AI Generated

DOI: 10.1007/s44408-026-00167-6

Keywords: PM2.5, natural ventilation, air quality sensors, WHO air quality guideline, indoor air quality, OpenAQ, low-cost sensors, temperature, relative humidity, Kuala Lumpur, Singapore, ventilation screening