A study on indoor air quality at daycare centers using IoT environmental sensors

Journal of Asian Architecture and Building Engineering, Год журнала: 2024, Номер unknown, С. 1 - 13

Опубликована: Окт. 20, 2024

Язык: Английский

---

Unseen crisis: Revealing the hidden health impact of indoor air pollution—A scoping review

Journal of Education and Health Promotion, Год журнала: 2024, Номер 13(1)

Опубликована: Окт. 1, 2024

Indoor air pollution presents a critical public health challenge, particularly in countries such as India, where millions are exposed to harmful pollutants within their homes and workplaces. This scoping review delves into the multifaceted impacts of indoor on outcomes, synthesizing evidence from various study designs geographical regions A was conducted. Drawing comprehensive search strategy, which yielded 320 records, wherein 120 PubMed, 108 Web Science, 92 SCOPUS. Ten studies were selected based predefined inclusion criteria, totaling sample size 37,43166 individuals . The synthesis findings reveals impact status. Respiratory symptoms illnesses found be prevalent among pollutants, with biomass fuel combustion posing high risk for chronic obstructive pulmonary disease (COPD) women. In addition, associated adverse pregnancy cardiovascular diseases, central nervous system impacts, cognitive impairment, developmental delays. Urgent action is needed reduce pollution, safeguard health, promote cleaner technologies healthier environments. Vulnerable populations, women, children, elderly, disproportionately affected by highlighting importance targeted interventions policies.

Язык: Английский

---

Design and Implementation of a Crowdsensing-Based Air Quality Monitoring Open and FAIR Data Infrastructure

Processes, Год журнала: 2023, Номер 11(7), С. 1881 - 1881

Опубликована: Июнь 23, 2023

This work reports on the development of a real-time vehicle sensor network (VSN) system and infrastructure devised to monitor particulate matter (PM) in urban areas within participatory paradigm. The approach is based use multiple vehicles where sensors, acquisition transmission devices are installed. PM values measured transmitted using standard mobile phone networks. Given large number platforms needed crowdsensing, sensors need be low-cost (LCS). sets limitations precision accuracy measurements that can mitigated statistical methods redundant data. Once data received, they automatically quality controlled, processed mapped geographically produce easy-to-understand visualizations made available almost real time through dedicated web portal. There, end users access current historic products. has been operational since 2021 collected over 50 billion measurements, highlighting several hotspots trends air pollution city Trieste (north-east Italy). study concludes (i) this perspective allows for drastically reduced costs considerably improves coverage measurements; (ii) an area approximately 100,000 square meters 200,000 inhabitants, quantity obtained with relatively low (5) public buses; (iii) small private cars, although less easy organize, very important provide infills buses not available; (iv) appropriate corrections LCS calculated applied reference taken high-quality standardized methods; (v) analyzing dispersion designated area, it possible highlight possibly associate them traffic directions. Crowdsensing open useful scientific community but also have great potential fostering environmental awareness adoption correct practices by general public.

Язык: Английский

---

Machine learning methods for low-cost pollen monitoring – Model optimisation and interpretability

The Science of The Total Environment, Год журнала: 2023, Номер 903, С. 165853 - 165853

Опубликована: Авг. 5, 2023

Pollen is a major issue globally, causing as much 40 % of the population to suffer from hay fever and other allergic conditions. Current techniques for monitoring pollen are either laborious slow, or expensive, thus alternative methods needed provide timely more localised information on airborne concentrations. We have demonstrated previously that low-cost Optical Particle Counter (OPC) sensors can be used estimate concentrations when machine learning process data learn relationships between OPC output conventionally measured This study demonstrates how methodical hyperparameter tuning employed significantly improve model performance. present results range models based tuned configurations trained predict Poaceae (Barnhart), Quercus (L.), Betula Pinus (L.) total The achieved here significant improvement we reported: average R2 scores at least doubled compared using previous parameter settings. Furthermore, employ explainable Artificial Intelligence (XAI) technique, SHAP, interpret understand each input features (i.e. particle sizes) affect estimated concentration type. In particular, found has strong positive correlation with particles optical diameter 1.7-2.3 μm, which distinguishes it types such may suggest type-specific subpollen in this size range. There further work done, especially training testing obtained across different environments evaluate extent generalisability. Nevertheless, potential method offer valuable insight gain what learned.

Язык: Английский

---

Construction and evaluation of hourly average indoor PM2.5 concentration prediction models based on multiple types of places

Frontiers in Public Health, Год журнала: 2023, Номер 11

Опубликована: Авг. 10, 2023

Background People usually spend most of their time indoors, so indoor fine particulate matter (PM 2.5 ) concentrations are crucial for refining individual PM exposure evaluation. The development concentration prediction models is essential the health risk assessment in epidemiological studies involving large populations. Methods In this study, based on monitoring data multiple types places, classical linear regression (MLR) method and random forest (RFR) algorithm machine learning were used to develop hourly average models. Indoor data, which included 11,712 records from five obtained by on-site monitoring. Moreover, potential predictor variable derived outdoor stations meteorological databases. A ten-fold cross-validation was conducted examine performance all proposed Results final variables incorporated MLR model concentration, type place, season, wind direction, surface speed, hour, precipitation, air pressure, relative humidity. results indicated that both constructed had good predictive performance, with determination coefficients (R 2 RFR 72.20 60.35%, respectively. Generally, better than (RFR developed using same as model, R = 71.86%). terms predictors, importance suggested speed important variables. Conclusion research, places first time. Both easily accessible indicators displayed promising domain outperformed result suggests application algorithms pollutant prediction.

Язык: Английский

---