Laboratory Evaluation of ARMIE, a Portable SPS30-Based Low-Cost Sensor Node for PM 2.5 Monitoring.

BACKGROUND: Low-cost particulate matter sensors have enabled new opportunities for exposure monitoring but require evaluation before application in epidemiological studies. This study assessed the performance of the SPS30 sensor integrated into the ARMIE portable monitoring sensor-node under controlled laboratory conditions.

METHODS: Sensors were co-located with two comparison instruments-the optical DustTrak photometer and the combined Scanning Mobility Particle Sizer (SMPS) and Aerodynamic Particle Sizer (APS)-across multiple aerosol sources, including candle burning, cooking, cigarette smoke, and clean air, under both regular and high-humidity conditions. Calibration performance was evaluated using leave-one-sensor-out and leave-one-source-out approaches.

RESULTS: The ARMIE node demonstrated strong agreement with the DustTrak ( r = 0.93-0.98) and maintained linear response characteristics across emission types. Calibration reduced mean errors and narrowed the limits of agreement. Agreement with the SMPS + APS was moderate ( r = 0.74-0.94) and characterized by systematic underestimation at higher concentrations.

CONCLUSIONS: Overall, the ARMIE node achieved high correlation with the DustTrak, demonstrating that low-cost optical sensors can reliably capture temporal variability in particle concentrations relative to mid-cost photometers.