Improving Air Filtration Based on PM2.5 Distribution

Dust concentrations in enclosed environments often exhibit considerable spatial variability, meaning that measurements taken at a single location may fail to capture localized areas of elevated exposure. To address this limitation, this study develops a spatial mapping approach for characterizing PM2.5 concentrations in a controlled test room and evaluating the influence of air filter placement on indoor particle distribution. A portable monitoring device integrating an ESP32 microcontroller and a PMS5003 particulate matter sensor is employed to collect measurements from 16 predefined sampling locations, with data automatically transmitted to Google Sheets. At each sampling point, a 30-second stabilization period is implemented before data collection, and consecutive valid readings are averaged to obtain representative PM2.5 concentrations. Spatial concentration distributions are subsequently reconstructed using Python-based inverse distance weighting (IDW) interpolation to visualize pollution patterns and identify potential exposure hotspots. The performance of different filter placement configurations is assessed through multiple quantitative indicators, including mean and maximum PM2.5 concentrations, hotspot frequency, and integrated pollutant concentration across identified hotspot areas. Comparative results and the optimal filter configuration will be established upon completion of the experimental measurements. Overall, this study proposes a preliminary, data-driven framework for spatially evaluating particulate matter distribution and optimizing air filter positioning. Further experimental validation under representative industrial conditions is necessary before extending the proposed approach to practical workshop environments.