Commercially available low-cost air quality sensors have low accuracy. The improved accuracy of low-cost PM_2.5 sensors allows the use of low-cost sensor systems to reasonably investigate PM_2.5 emissions from industrial activities or to accurately estimate individual exposure to PM_2.5. In this work, we developed a new PM_2.5 calibration model (HybridLSTM) by combining a deep neural network (DNN) optimized in calibration problems and a long short-term memory (LSTM) neural network optimized in time-dependent characteristics to improve the performance of conventional calibration algorithms of low-cost PM sensors. The PM_2.5 concentrations, temperature and humidity by low-cost sensors and gravimetric-based PM_2.5 measuring instrument were sampled for a sufficiently long time. The proposed model was compared with benchmarks (multiple linear regression model (MLR), DNN model) and low-cost sensor results. The gravimetric measurements were used as reference data to evaluate sensor accuracy. For root-mean-square error (RMSE) for PM_2.5 concentrations, the proposed model reduced 41–60% of error when compared with the raw data of low-cost sensors, reduced 30–51% of error when compared with the MLR model and reduced 8–40% of error when compared with the MLR model. R² of HybridLSTM, DNN, MLR and raw data were 93, 90, 80 and 59%, respectively. HybridLSTM showed the state-of-the-art calibration performance for a low-cost PM sensor. In other words, the proposed ML model has state-of-the-art calibration performance among the tested calibration algorithms.

The evaporation of chemicals in applied chemical products such as cleaning products and paint has been evaluated using the evaporation mode of the ConsExpo model. However, it remains controversial whether the ConsExpo model can be used for non-applied chemical products such as air fresheners, because the ConsExpo model assumes that the mass of the non-applied chemical products does not change during the time of use. If most of the materials in the product are volatile, the product mass can decrease. To explain the effect of a change in product mass, the ‘Chemical Product Evaporation Model (CPEM)’ was developed. This study demonstrated that the product mass decreases linearly when the surface area of the product is invariant, theoretically and experimentally. It was found that the ConsExpo evaporation model can be applied to products in which the other materials do not evaporate, and the CPEM can be applied to products in which the other materials are volatile. If the target substance in a product evaporates completely before the exhaustion time, the average concentration of a target substance in the air can be estimated simply from its initial concentration in the product and the product mass reduction rate. Otherwise, we recommend using the CPEM.