Particulate matter represents a significant health and safety concern. PM2.5 and PM10 exposure at workplace can cause respiratory diseases, cardiovascular disorders, and long-term occupational illness. Many countries’ occupational health and safety regulations define permissible concentration exposure limits and recommend continuous monitoring. This research aims to develop a predictive model that identifies PM2.5 and PM10 exceedances, enabling timely warnings to wheel loaders’ operators and helping to prevent exposure to its elevated levels. A multilayer feedforward neural network for PM2.5 prediction and a long short-term memory network for PM10 exceeding prediction were trained using 469 on-site measurements, from sensors placed in the cabins of 8 wheel loaders on different open pit mines. Results showed that both models have a significant level of reliability and repeatability with an accuracy greater than 88%. A further research avenue proposal is to enlarge the training sample and to incorporate alert system in the wheel loaders’ cabin.
Janev et al. (Thu,) studied this question.