MoreAir: A Low-Cost Urban Air Pollution Monitoring System

Real-time AIoT platform for monitoring and prediction of air quality in Southwestern Morocco

Urbanization and industrialization have led to a significant increase in air pollution, posing a severe environmental and public health threat. Accurate forecasting of air quality is crucial for policymakers to implement effective interventions. This study presents a novel AIoT platform specifically designed for PM2.5 monitoring in Southwestern Morocco. The platform utilizes low-cost sensors to collect air quality data, transmitted via WiFi/3G for analysis and prediction on a central server. We focused on identifying optimal features for PM2.5 prediction using Minimum Redundancy Maximum Relevance (mRMR) and LightGBM Recursive Feature Elimination (LightGBM-RFE) techniques. Furthermore, Bayesian optimization was employed to fine-tune hyperparameters of popular machine learning models for the most accurate PM2.5 concentration forecasts. Model performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R2). Our results demonstrate that the LightGBM model achieved superior performance in PM2.5 prediction, with a significant reduction in RMSE compared to other evaluated models. This study highlights the potential of AIoT platforms coupled with advanced feature selection and hyperparameter optimization for effective air quality monitoring and forecasting.

Calibration of a Low-Cost Methane Sensor Using Machine Learning

In order to combat greenhouse gas emissions, the sources of these emissions must be understood. Environmental monitoring using low-cost wireless devices is one method of measuring emissions in crucial but remote settings, such as peatlands. The Figaro NGM2611-E13 is a low-cost methane detection module based around the TGS2611-E00 sensor. The manufacturer provides sensitivity characteristics for methane concentrations above 300 ppm, but lower concentrations are typical in outdoor settings. This study investigates the potential to calibrate these sensors for lower methane concentrations using machine learning. Models of varying complexity, accounting for temperature and humidity variations, were trained on over 50,000 calibration datapoints, spanning 0-200 ppm methane, 5-30 °C and 40-80% relative humidity. Interaction terms were shown to improve model performance. The final selected model achieved a root-mean-square error of 5.1 ppm and an R2 of 0.997, demonstrating the potential for the NGM2611-E13 sensor to measure methane concentrations below 200 ppm.

Low-cost urban carbon monitoring network and implications for china: a comprehensive review

The development and renewal of gas sensor technology have enabled more and more low-cost gas sensors to form a carbon monitoring network to meet the requirements of the city. In the context of China's commitment to achieving the "double carbon" target by 2060, this paper reviews the principles of four standard gas sensors and the application of several low-cost sensors in urban carbon monitoring networks, with the aim of providing a practical reference for the future deployment of carbon monitoring networks in Chinese cities. Moreover, the types, prices, and deployment of the sensors used in each project are summarized. Based on this review, non-dispersive infrared sensors have the best performance among the sensors and are commonly used in many cities. Lots of urban climate networks in cities were summarized by many reviews in the literature, but only a few sensors were studied, and they did not consider carbon dioxide (CO2) sensors. This review focuses on the dense CO2 urban monitoring network, and some case studies are also discussed, such as Seoul and San Francisco. To address the issue of how to better ensure the balance between cost and accuracy in the deployment of sensor networks, this paper proposes a method of simultaneously deploying medium-precision and high-precision fixed sensors and mobile sensors to form an urban carbon monitoring network. Finally, the prospects and recommendations, such as different ways to mitigate CO2 and develop an entire carbon monitoring system for future urban carbon monitoring in China, are also presented.

Air pollution forecasting based on wireless communications: review

The development of contemporary artificial intelligence (AI) methods such as artificial neural networks (ANNs) has given researchers around the world new opportunities to address climate change and air quality issues. The small size, low cost, and low power consumption of sensors can facilitate obtaining the values of polluting gases in the atmosphere. However, several problems with using air pollution technique relate to various effects such as sensing accuracy, sensor drifts, and sluggish reactions to changes in pollution levels. Recently, machine learning has made it feasible to build a more intelligent, context-aware system that can anticipate events and monitor present conditions. This paper focuses on the use of environment sensors for detecting air pollution based on several types of wireless protocols, including Wi-Fi, Bluetooth, ZigBee, LoRa, Global Positioning System (GPS), and 4G/5G. Furthermore, it classifies previous published articles on the topic according to the wireless protocol and compared in terms of several performance metrics such as the adopted air pollution sensors, hardware platform, adopted algorithm, power consumption or power savings, and sensing accuracy. In addition, this work highlights the challenges and limitations facing drones during their mission for detecting air pollution. As a result, we suggest to build and implement at base station an intelligent system based on backpropagation (BP) neural networks, which provides flexibility to track and predict the true values of polluting gases in the atmosphere to overcome the above problems. Finally, this work addresses the advantages of using drones in the air pollution field.

Low-Cost Sensor for Lycopene Content Measurement in Tomato Based on Raspberry Pi 4

Measuring lycopene in tomatoes is fundamental to the agrifood industry because of its health benefits. It is one of the leading quality criteria for consuming this fruit. Traditionally, the amount determination of this carotenoid is performed using the high-performance liquid chromatography (HPLC) technique. This is a very reliable and accurate method, but it has several disadvantages, such as long analysis time, high cost, and destruction of the sample. In this sense, this work proposes a low-cost sensor that correlates the lycopene content in tomato with the color present in its epicarp. A Raspberry Pi 4 programmed with Python language was used to develop the lycopene prediction model. Various regression models were evaluated using neural networks, fuzzy logic, and linear regression. The best model was the fuzzy nonlinear regression as the RGB input, with a correlation of R² = 0.99 and a mean error of 1.9 × 10^-5. This work was able to demonstrate that it is possible to determine the lycopene content using a digital camera and a low-cost integrated system in a non-invasive way.

Calibration Assessment of Low-Cost Carbon Dioxide Sensors Using the Extremely Randomized Trees Algorithm

As the monitoring of carbon dioxide is an important proxy to estimate the air quality of indoor and outdoor environments, it is essential to obtain trustful data from CO₂ sensors. However, the use of widely available low-cost sensors may imply lower data quality, especially regarding accuracy. This paper proposes a new approach for enhancing the accuracy of low-cost CO₂ sensors using an extremely randomized trees algorithm. It also reports the results obtained from experimental data collected from sensors that were exposed to both indoor and outdoor environments. The indoor experimental set was composed of two metal oxide semiconductors (MOS) and two non-dispersive infrared (NDIR) sensors next to a reference sensor for carbon dioxide and independent sensors for air temperature and relative humidity. The outdoor experimental exposure analysis was performed using a third-party dataset which fit into our goals: the work consisted of fourteen stations using low-cost NDIR sensors geographically spread around reference stations. One calibration model was trained for each sensor unit separately, and, in the indoor experiment, it managed to reduce the mean absolute error (MAE) of NDIR sensors by up to 90%, reach very good linearity with MOS sensors in the indoor experiment (r² value of 0.994), and reduce the MAE by up to 98% in the outdoor dataset. We have found in the outdoor dataset analysis that the exposure time of the sensor itself may be considered by the algorithm to achieve better accuracy. We also observed that even a relatively small amount of data may provide enough information to perform a useful calibration if they contain enough data variety. We conclude that the proper use of machine learning algorithms on sensor readings can be very effective to obtain higher data quality from low-cost gas sensors either indoors or outdoors, regardless of the sensor technology.

Assessment of the impact of sensor error on the representativeness of population exposure to urban air pollutants

For the monitoring of urban air pollution, smart sensors are often seen as a welcome addition to fixed-site monitoring (FSM) networks. Due to price and simple installation, increases in spatial representation are thought to be achieved by large numbers of these sensors, however, a number of sensor errors have been identified. Based on a high-resolution modelling system, up to 400 pseudo smart sensors were perturbated with the aim of simulating common sensor errors and added to the existing FSM network in Hong Kong, resulting in 1200 pseudo networks for PM_2.5 and 1040 pseudo networks for NO₂. For each pseudo network, population-weighted area representativeness (PWAR) was calculated based on similarity frequency. For PM_2.5, improvements (up to 16%) to the high baseline representativeness (PWAR = 0.74) were achievable only by the addition of high-quality sensors and favourable environmental conditions. The baseline FSM network represents NO₂ less well (PWAR = 0.52), as local emissions in the study domain resulted in high spatial pollution variation. Due to higher levels of pollution (population-weighted average 37.3 ppb) in comparison to sensor error ranges, smart sensors of a wider quality range were able to improve network representativeness (up to 42%). Marginal representativeness increases were found to exponentially decrease with existing sensor number. The quality and maintenance of added sensors had a stronger effect on overall network representativeness than the number of sensors added. Often, a small number of added sensors of a higher quality class led to larger improvements than hundreds of lower-class sensors. Whereas smart sensor performance and maintenance are important prerequisites particularly for developed cities where pollutant concentration is low and there is an existing FSM network, our study shows that for places with high pollutant variability and concentration such as encountered in some developing countries, smart sensors will provide benefits for understanding population exposure.

Development and Performance Evaluation of a Low-Cost Portable PM2.5 Monitor for Mobile Deployment

The concentration of fine particulate matter (PM2.5) is known to vary spatially across a city landscape. Current networks of regulatory air quality monitoring are too sparse to capture these intra-city variations. In this study, we developed a low-cost (60 USD) portable PM2.5 monitor called Smart-P, for use on bicycles, with the goal of mapping street-level variations in PM2.5 concentration. The Smart-P is compact in size (85 × 85 × 42 mm) and light in weight (147 g). Data communication and geolocation are achieved with the cyclist’s smartphone with the help of a user-friendly app. Good agreement was observed between the Smart-P monitors and a regulatory-grade monitor (mean bias error: −3.0 to 1.5 μg m−3 for the four monitors tested) in ambient conditions with relative humidity ranging from 38 to 100%. Monitor performance decreased in humidity > 70% condition. The measurement precision, represented as coefficient of variation, was 6 to 9% in stationary mode and 6% in biking mode across the four tested monitors. Street tests in a city with low background PM2.5 concentrations (8 to 9 μg m−3) and in two cities with high background concentrations (41 to 74 μg m−3) showed that the Smart-P was capable of observing local emission hotspots and that its measurement was not sensitive to bicycle speed. The low-cost and user-friendly nature are two features that make the Smart-P a good choice for empowering citizen scientists to participate in local air quality monitoring.

An Exploration of Features Impacting Respiratory Diseases in Urban Areas

Air pollution exposure has become ubiquitous and is increasingly detrimental to human health. Small Particulate matter (PM) is one of the most harmful forms of air pollution. It can easily infiltrate the lungs and trigger several respiratory diseases, especially in vulnerable populations such as children and elderly people. In this work, we start by leveraging a retrospective study of 416 children suffering from respiratory diseases. The study revealed that asthma prevalence was the most common among several respiratory diseases, and that most patients suffering from those diseases live in areas of high traffic, noise, and greenness. This paved the way to the construction of the MOREAIR dataset by combining feature abstraction and micro-level scale data collection. Unlike existing data sets, MOREAIR is rich in context-specific components, as it includes 52 temporal or geographical features, in addition to air-quality measurements. The use of Random Forest uncovered the most important features for the understanding of air-quality distribution in Moroccan urban areas. By linking the medical data and the MOREAIR dataset, we observed that the patients included in the medical study come mostly from neighborhoods that are characterized by either high average or high variations of pollution levels.

Socioeconomic Disparities of Low-Cost Air Quality Sensors in California, 2017-2020

Objectives. To (1) examine the disparity in availability of PurpleAir low-cost air quality sensors in California based on neighborhood socioeconomic status (SES) and exposure to fine particulate matter smaller than 2.5 micrometers (PM2.5), (2) investigate the temporal trend of sensor distribution and operation, and (3) identify priority communities for future sensor distribution. Methods. We obtained census tract-level SES variables and PM2.5 concentrations from the CalEnviroScreen4.0 data set. We obtained real-time PurpleAir sensor data (July 2017-September 2020) to examine sensor distribution and operation. We conducted spatial and temporal analyses at the census tract level to investigate neighborhood SES and PM2.5 concentrations in relation to sensor distribution and operation. Results. The spatial coverage and the number of PurpleAir sensors increased significantly in California. Fewer sensors were distributed in census tracts with lower SES, higher PM2.5, and higher proportions of racial/ethnic minority populations. Furthermore, a large proportion of existing sensors were not in operation at a given time, especially in disadvantaged communities. Conclusions. Disadvantaged communities should be given access to low-cost sensors to fill in spatial gaps of air quality monitoring and address environmental justice concerns. Sensor purchasing and deployment must be paired with regular maintenance to ensure their reliable performance. (Am J Public Health. 2022;112(3):434-442. https://doi.org/10.2105/AJPH.2021.306603).

Environmental Monitoring of the Littoral Zone of Lake Baikal Using a Network of Automatic Hydro-Meteorological Stations: Development and Trial Run

An automatic hydro-meteorological station (AHMS) was designed to monitor the littoral zone of Lake Baikal in areas with high anthropogenic pressure. The developed AHMS was installed near the Bolshiye Koty settlement (southern basin). This AHMS is the first experience focused on obtaining the necessary competencies for the development of a monitoring network of the Baikal natural territory. To increase the flexibility of adjustment and repeatability, we developed AHMS as a low-cost modular system. AHMS is equipped with a weather station and sensors measuring water temperature, pH, dissolved oxygen, redox potential, conductivity, chlorophyll-a, and turbidity. This article describes the main AHMS functions (hardware and software) and measures taken to ensure data quality control. We present the results of the first two periods of its operation. The data acquired during this periods have demonstrated that, to obtain accurate measurements and to detect and correct errors that were mainly due to biofouling of the sensors and calibration bias, a correlation between AHMS and laboratory studies is necessary for parameters such as pH and chlorophyll-a. The gained experience should become the basis for the further development of the monitoring network of the Baikal natural territory.

Machine Learning for Predicting the Risk for Childhood Asthma Using Prenatal, Perinatal, Postnatal and Environmental Factors

The prevalence rate for childhood asthma and its associated risk factors vary significantly across countries and regions. In the case of Morocco, the scarcity of available medical data makes scientific research on diseases such as asthma very challenging. In this paper, we build machine learning models to predict the occurrence of childhood asthma using data from a prospective study of 202 children with and without asthma. The association between different factors and asthma diagnosis is first assessed using a Chi-squared test. Then, predictive models such as logistic regression analysis, decision trees, random forest and support vector machine are used to explore the relationship between childhood asthma and the various risk factors. First, data were pre-processed using a Chi-squared feature selection, 19 out of the 36 factors were found to be significantly associated (p-value < 0.05) with childhood asthma; these include: history of atopic diseases in the family, presence of mites, cold air, strong odors and mold in the child's environment, mode of birth, breastfeeding and early life habits and exposures. For asthma prediction, random forest yielded the best predictive performance (accuracy = 84.9%), followed by logistic regression (accuracy = 82.57%), support vector machine (accuracy = 82.5%) and decision trees (accuracy = 75.19%). The decision tree model has the advantage of being easily interpreted. This study identified important maternal and prenatal risk factors for childhood asthma, the majority of which are avoidable. Appropriate steps are needed to raise awareness about the prenatal risk factors.

Current advances and future challenges of AIoT applications in particulate matters (PM) monitoring and control

Air pollution is at the center of pollution-control discussion due to the significant adverse health effects on individuals and the environment. Research has shown the association between unsafe environments and different sizes of particulate matter (PM), highlighting the importance of pollutant monitoring to mitigate its detrimental effect. By monitoring air quality with low-cost monitoring devices that collect massive observations, such as Air Box, a comprehensive collection of ground-level PM concentration is plausible due to the simplicity and low-cost, propelling applications in agriculture, aquaculture, and air quality, water resources, and disaster prevention. This paper aims to view IoT-based systems with low-cost microsensors at the sensor, network, and application levels, along with machine learning algorithms that improve sensor networks' precision, providing better resolution. From the analysis at the three levels, we analyze current PM monitoring methods, including the use of sensors when collecting PM concentrations, demonstrate the use of IoT-based systems in PM monitoring and its challenges, and finally present the integration of AI and IoT (AIoT) in PM monitoring, indoor air quality control, and future directions. In addition, the inclusion of Taiwan as a site analysis was illustrated to show an example of AIoT in PM-control policy-making potential directions.

An Approximation for Metal-Oxide Sensor Calibration for Air Quality Monitoring Using Multivariable Statistical Analysis

Good air quality is essential for both human beings and the environment in general. The three most harmful air pollutants are nitrogen dioxide (NO₂), ozone (O₃) and particulate matter. Due to the high cost of monitoring stations, few examples of this type of infrastructure exist, and the use of low-cost sensors could help in air quality monitoring. The cost of metal-oxide sensors (MOS) is usually below EUR 10 and they maintain small dimensions, but their use in air quality monitoring is only valid through an exhaustive calibration process and subsequent precision analysis. We present an on-field calibration technique, based on the least squares method, to fit regression models for low-cost MOS sensors, one that has two main advantages: it can be easily applied by non-expert operators, and it can be used even with only a small amount of calibration data. In addition, the proposed method is adaptive, and the calibration can be refined as more data becomes available. We apply and evaluate the technique with a real dataset from a particular area in the south of Spain (Granada city). The evaluation results show that, despite the simplicity of the technique and the low quantity of data, the accuracy obtained with the low-cost MOS sensors is high enough to be used for air quality monitoring.

Remote Tracking Gas Molecular via the Standalone-Like Nanosensor-Based Tele-Monitoring System

Highlights

A standalone-like smart device that can remotely track the variation of air pollutants in a power-saving way is created; Metal–organic framework-derived hollow polyhedral ZnO was successfully synthesized, allowing the created smart device to be highly selective and to sensitively track the variation of NO2 concentration; A novel photoluminescence-enhanced Li-Fi telecommunication technique is proposed, offering the created smart device with the capability of long distance wireless communication.

Abstract

Remote tracking the variation of air quality in an effective way will be highly helpful to decrease the health risk of human short- and long-term exposures to air pollution. However, high power consumption and poor sensing performance remain the concerned issues, thereby limiting the scale-up in deploying air quality tracking networks. Herein, we report a standalone-like smart device that can remotely track the variation of air pollutants in a power-saving way. Brevity, the created smart device demonstrated satisfactory selectivity (against six kinds of representative exhaust gases or air pollutants), desirable response magnitude (164–100 ppm), and acceptable response/recovery rate (52.0/50.5 s), as well as linear response relationship to NO2. After aging for 2 weeks, the created device exhibited relatively stable sensing performance more than 3 months. Moreover, a photoluminescence-enhanced light fidelity (Li-Fi) telecommunication technique is proposed and the Li-Fi communication distance is significantly extended. Conclusively, our reported standalone-like smart device would sever as a powerful sensing platform to construct high-performance and low-power consumption air quality wireless sensor networks and to prevent air pollutant-induced diseases via a more effective and low-cost approach.

Electronic supplementary material

The online version of this article (10.1007/s40820-020-00551-w) contains supplementary material, which is available to authorized users.