Challenges and Opportunities in Monitoring Indoor Air Quality with Low-Cost Sensors

Low-cost air quality sensors offer promising, cost-effective solutions for monitoring indoor air quality (IAQ). However, their utility is often constrained by challenges in accuracy and data reliability, including those related to inadequate calibration. This is particularly important considering the large variabilities in indoor environmental conditions, in terms of temperature, humidity, and PM_2.5 levels. To address this, we developed an experimental chamber using a container to calibrate two types of low-cost particulate matter (PM) sensors against a reference-grade instrument (Fidas 200) using machine learning methods. Sensors from two brands were tested in controlled conditions simulating common indoor pollutant sources such as cooking, smoking, and incense burning. Calibration revealed clear performance variability between sensor brands, with sensors underestimating or overestimating pollutant concentrations at different levels. Sensor correction using machine learning greatly improved sensor accuracy and data reliability. Calibrated sensors will be deployed to monitor PM concentrations (PM_2.5 and PM₁₀), temperature, relative humidity, and carbon dioxide (CO₂) levels continuously over two years in retrofitted and control homes, capturing pre- and post-retrofit IAQ changes. Future recalibration is planned to evaluate long-term sensor drift. Our preliminary findings highlight the critical role of rigorous calibration in ensuring reliable IAQ monitoring using low-cost sensors. This study provides valuable insights into the practical applications and limitations of such sensors in retrofitted environments.