BUSCADOR RECOLECTA

Growing progress in sensor technology has constantly expanded the number and range of low-cost, small, and portable sensors on the market, increasing the number and type of physical phenomena that can be measured with wirelessly connected sensors. Large-scale deployments of wireless sensor networks (WSN) involving hundreds or thousands of devices and limited budgets often constrain the choice of sensing hardware, which generally has reduced accuracy, precision, and reliability. Therefore, it is challenging to achieve good data quality and maintain error-free measurements during the whole system lifetime. Self-calibration or recalibration in ad hoc sensor networks to preserve data quality is essential, yet challenging, for several reasons, such as the existence of random noise and the absence of suitable general models. Calibration performed in the field, without accurate and controlled instrumentation, is said to be in an uncontrolled environment. This paper provides current and fundamental self-calibration approaches and models for wireless sensor networks in uncontrolled environments., Peer Reviewed

The remarkable advances in sensing and communication technologies have introduced increasingly low-cost, smart and portable sensors that can be embedded everywhere and play an important role in environmental sensing applications such as air quality monitoring. These user-friendly wireless sensor platforms enable assessment of human exposure to air pollution through observations at high spatial resolution in near-realtime, thus providing new opportunities to simultaneously enhance existing monitoring systems, as well as engage citizens in active environmental monitoring. However, data quality from such platforms is a concern since sensing hardware of such devices is generally characterized by a reduced accuracy, precision, and reliability. Achieving good data quality and maintaining error free measurements during the whole system lifetime is challenging. Over time, sensors become subject to several sources of unknown and uncontrollable faulty data which comprise the accuracy of the measurements and yield observations far from the expected values. This paper investigates calibration of low-cost air quality sensors in a real sensor network deployment. The approach leverages on the availability of sensor arrays in a wireless node to estimate parameters that minimize the calibration error using fusion of data from multiple sensors. The obtained results were encouraging and show the effectiveness of the approach compared to a single sensor calibration., Peer Reviewed

© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., This article investigates the calibration of low-cost sensors for air pollution. The sensors were deployed on three Internet of Things (IoT) platforms in Spain, Austria, and Italy during the summers of 2017, 2018, and 2019. One of the biggest challenges in the operation of an IoT platform, which has a great impact on the quality of the reported pollution values, is the calibration of the sensors in an uncontrolled environment. This calibration is performed using arrays of sensors that measure cross sensitivities and therefore compensate for both interfering contaminants and environmental conditions. This article investigates how the fusion of data taken by sensor arrays can improve the calibration process. In particular, calibration with sensor arrays, multisensor data fusion calibration with weighted averages, and multisensor data fusion calibration with machine learning models are compared. Calibration is evaluated by combining data from various sensors with linear and nonlinear regression models., This work was supported in part by the NationalSpanish funding under Grant TIN2016-78473-C3-1-R, in part by RegionalProject under Grant 2017SGR-990, in part by EU H2020 CAPTOR Project, inpart by AGAUR SGR44, in part by the Agencia Estatal de Investigación underGrant CGL2017-82093-ERC, and in part by the Spanish Ministry of Economy,Industry and Competitiveness under Grant EUIN2017-85799, Peer Reviewed

New advances in sensor technologies and communications in wireless sensor networks have favored the introduction of low-cost sensors for monitoring air quality applications. In this article, we present the results of the European project H2020 CAPTOR, where three testbeds with sensors were deployed to capture tropospheric ozone concentrations. One of the biggest challenges was the calibration of the sensors, as the manufacturer provides them without calibrating. Throughout the paper, we show how short-term calibration using multiple linear regression produces good calibrated data, but instead produces biases in the calculated long-term concentrations. To mitigate the bias, we propose a linear correction based on Kriging estimation of the mean and standard deviation of the long-term ozone concentrations, thus correcting the bias presented by the sensors., Peer Reviewed

Tropospheric ozone (O3) is an environmental pollutant of growing concern, especially in suburban and rural areas where the density of air quality monitoring stations is not high. In this type of areas citizen science strategies can be useful tools for awareness raising, but sensor technologies must be validated before sensor data are communicated to the public. In this work, the performance under field conditions of two custom-made types of ozone sensing devices, based on metal-oxide and electrochemical sensors, was tested. A large array of 132 metal-oxide (Sensortech MICS 2614) and 11 electrochemical (Alphasense) ozone sensors, built into 44 sensing devices, was co-located at reference stations in Italy (4 stations) and Spain (5). Mean R2 between sensor and reference data was 0.88 (0.78–0.96) and 0.89 (0.73–0.96) for Captor (metal-oxide) and Raptor (electrochemical) nodes. The metal-oxide sensors showed an upper limit (approximately 170 µg/m3) implying that these sensors may be useful to communicate mean ozone concentrations but not peak episodes. The uncertainty of the nodes was 10% between 100 and 150 µg/m3 and 20% between 150 and 200 µg/m3, for Captors, and 10% for >100 µg/m3 for Raptors. Operating both types of nodes up to 5 months did not evidence any clear influence of drifts. The use of these sensors in citizen science can be a useful tool for awareness raising. However, significant data processing efforts are required to ensure high data quality, and thus machine learning strategies are advisable. Relative uncertainties should always be reported when communicating ozone concentration data from sensing nodes., Peer Reviewed

© 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes,creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., This paper shows the result of the calibration process of an Internet of Things platform for the measurement of tropospheric ozone (O 3 ). This platform, formed by 60 nodes, deployed in Italy, Spain, and Austria, consisted of 140 metal–oxide O 3 sensors, 25 electro-chemical O 3 sensors, 25 electro-chemical NO 2 sensors, and 60 temperature and relative humidity sensors. As ozone is a seasonal pollutant, which appears in summer in Europe, the biggest challenge is to calibrate the sensors in a short period of time. In this paper, we compare four calibration methods in the presence of a large dataset for model training and we also study the impact of a limited training dataset on the long-range predictions. We show that the difficulty in calibrating these sensor technologies in a real deployment is mainly due to the bias produced by the different environmental conditions found in the prediction with respect to those found in the data training phase., Peer Reviewed

The H2020 CAPTOR project deployed three testbeds in Spain, Italy and Austria with low-cost sensors for the measurement of tropospheric ozone (O3). The aim of the H2020 CAPTOR project was to raise public awareness in a project focused on citizen science. Each testbed was supported by an NGO in charge of deciding how to raise citizen awareness according to the needs of each country. The data presented in this document correspond to the raw data captured by the sensor nodes in the Spanish testbed using SGX Sensortech MICS 2614 metal-oxide sensors. The Spanish testbed consisted of the deployment of twenty-five nodes. Each sensor node included four SGX Sensortech MICS 2614 ozone sensors, one temperature sensor and one relative humidity sensor. Each node underwent a calibration process by co-locating the node at an EU reference air quality monitoring station, followed by a deployment in a sub-urban or rural area in Catalonia, Spain. All nodes spent two to three weeks co-located at a reference station in Barcelona, Spain (urban area), followed by two to three weeks co-located at three sub-urban reference stations near the final deployment site. The nodes were then deployed in volunteers' homes for about two months and, finally, the nodes were co-located again at the sub-urban reference stations for two weeks for final calibration and assessment of potential drifts. All data presented in this paper are raw data taken by the sensors that can be used for scientific purposes such as calibration studies using machine learning algorithms, or once the concentration values of the nodes are obtained, they can be used to create tropospheric ozone pollution maps with heterogeneous data sources (reference stations and low-cost sensors)., This work was funded by H2020 project CAPTOR, and supported by National Spanish project PID2019–107910RB-I00, and regional project 2017SGR-990, and with the support of Secretaria d’Universitats i Recerca de la Generalitat de Catalunya i del Fons Social Europeu. The authors gratefully acknowledge the collaboration of the staff at the Department of the Environment of the Generalitat de Catalunya who kindly provided support for the deployment of the sensing nodes at reference stations for calibrating the sensors and provided access to the reference data., Peer Reviewed

Tropospheric ozone (O3) is an environmental pollutant of growing concern, especially in suburban and rural areas where the density of air quality monitoring stations is not high. In this type of areas citizen science strategies can be useful tools for awareness raising, but sensor technologies must be validated before sensor data are communicated to the public. In this work, the performance under field conditions of two custom-made types of ozone sensing devices, based on metal-oxide and electrochemical sensors, was tested. A large array of 132 metal-oxide (Sensortech MICS 2614) and 11 electrochemical (Alphasense) ozone sensors, built into 44 sensing devices, was co-located at reference stations in Italy (4 stations) and Spain (5). Mean R2 between sensor and reference data was 0.88 (0.78–0.96) and 0.89 (0.73–0.96) for Captor (metal-oxide) and Raptor (electrochemical) nodes. The metal-oxide sensors showed an upper limit (approximately 170 μg/m3) implying that these sensors may be useful to communicate mean ozone concentrations but not peak episodes. The uncertainty of the nodes was 10% between 100 and 150 μg/m3 and 20% between 150 and 200 μg/m3, for Captors, and 10% for >100 μg/m3 for Raptors. Operating both types of nodes up to 5 months did not evidence any clear influence of drifts. The use of these sensors in citizen science can be a useful tool for awareness raising. However, significant data processing efforts are required to ensure high data quality, and thus machine learning strategies are advisable. Relative uncertainties should always be reported when communicating ozone concentration data from sensing nodes. © 2018 The Authors, The authors would like to thank the anonymous reviewers, whose comments improved the manuscript significantly. This work was funded by H2020 project CAPTOR. The authors gratefully acknowledge the collaboration of the staff at the Department of the Environment of the Generalitat de Catalunya and at ARPA Lombardia (Agenzia Regionale per l'Ambiente), who kindly provided support for the deployment of the sensing nodes at reference stations and provided access to the reference data. Support is also acknowledged from AGAUR SGR44, the Agencia Estatal de Investigación (AEI; CGL2017-82093-ERC ), and the Spanish Ministry of Economy, Industry and Competitiveness ( EUIN2017-85799 ). Appendix A, Peer reviewed

New advances in sensor technologies and communications in wireless sensor networks have favored the introduction of low-cost sensors for monitoring air quality applications. In this article, we present the results of the European project H2020 CAPTOR, where three testbeds with sensors were deployed to capture tropospheric ozone concentrations. One of the biggest challenges was the calibration of the sensors, as the manufacturer provides them without calibrating. Throughout the paper, we show how short-term calibration using multiple linear regression produces good calibrated data, but instead produces biases in the calculated long-term concentrations. To mitigate the bias, we propose a linear correction based on Kriging estimation of the mean and standard deviation of the long-term ozone concentrations, thus correcting the bias presented by the sensors. © 2019, MDPI AG. All rights reserved., Funding text #1
Universitat Politecnica de Catalunya (UPC), UPC Campus Nord, 08034 Barcelona, Spain; pauferrercid12@gmail.com (P.F.-C.); jorge@ac.upc.edu (J.G.-V.) Institute of Environmental Assessment and Water Research, Spanish National Research Council (IDAEA-CSIC), 08034 Barcelona, Spain; anna.ripoll@idaea.csic.es (A.R.); mar.viana@idaea.csic.es (M.V.) Correspondence: joseb@ac.upc.edu; Tel.:+34-93-405-4051

Funding text #2
Funding: This work is supported by the National Spanish funding TIN2016-78473-C3-1-R, regional Spanish funding AGAUR 2017SGR-990, 2017-SGR-44, the Agencia Estatal de Investigación (AEI; CGL2017-82093-ERC), and the Spanish Ministry of Economy, Industry and Competitiveness (EUIN2017-85799) and European H2020 CAPTOR project., Peer reviewed

This article investigates the calibration of low-cost sensors for air pollution. The sensors were deployed on three Internet of Things (IoT) platforms in Spain, Austria, and Italy during the summers of 2017, 2018, and 2019. One of the biggest challenges in the operation of an IoT platform, which has a great impact on the quality of the reported pollution values, is the calibration of the sensors in an uncontrolled environment. This calibration is performed using arrays of sensors that measure cross sensitivities and therefore compensate for both interfering contaminants and environmental conditions. This article investigates how the fusion of data taken by sensor arrays can improve the calibration process. In particular, calibration with sensor arrays, multisensor data fusion calibration with weighted averages, and multisensor data fusion calibration with machine learning models are compared. Calibration is evaluated by combining data from various sensors with linear and nonlinear regression models., National Spanish funding; Regional Project; EU H2020 CAPTOR Project; AGAUR SGR44; 10.13039/501100011033-Agencia Estatal de Investigación; Spanish Ministry of Economy, Industry and Competitiveness;, Peer reviewed

The data folder is includes the five different data sets used in the paper along with a metadata file, where the different features are explained., This dataset contains python scripts to calibrate tropospheric ozone sensor data obtained in the H2020 Captor project using sensor fusion techniques. Four different models are implemented; Multiple Linear Regression (MLR), K-Nearest Neighbors (KNN),Random Forest(RF) and Support Vector Regression (SVR). The methodology consists of first applying the PLS procedure to derive orthogonal components (to avoid multicollinearity problems). Afterwards, the components are used as features in the machine learning algorithms, so the models are trained. The scripts available in this repository have been used in the elaboration of the paper: "Multi-sensor data fusion calibration in IoT air pollution platforms" submitted to the IEEE Internet of Things journal., National Spanish funding; Regional Project; EU H2020 CAPTOR Project; AGAUR SGR44; 10.13039/501100011033-Agencia Estatal de Investigación; Spanish Ministry of Economy, Industry and Competitiveness, Peer reviewed

The H2020 CAPTOR project deployed three testbeds in Spain, Italy and Austria with low-cost sensors for the measurement of tropospheric ozone (O3). The aim of the H2020 CAPTOR project was to raise public awareness in a project focused on citizen science. Each testbed was supported by an NGO in charge of deciding how to raise citizen awareness according to the needs of each country. The data presented here correspond to the raw data captured by the sensor nodes in the Spanish testbed using SGX Sensortech MICS 2614 metal-oxide sensors. The Spanish testbed consisted of the deployment of twenty-five nodes. Each sensor node included four SGX Sensortech MICS 2614 ozone sensors, one temperature sensor and one relative humidity sensor. Each node underwent a calibration process by co-locating the node at a reference station, followed by a deployment in a non-urban area in Catalonia, Spain. All nodes spent two to three weeks co-located at a reference station in Barcelona, Spain (urban area), followed by two to three weeks co-located at three non-urban reference stations near the final deployment site. The nodes were then deployed in volunteers' homes for about two months and, finally, the nodes were co-located again at the non-urban reference stations for two weeks. All data presented in this repository are raw data taken by the sensors that can be used for scientific purposes such as calibration studies using machine learning algorithms, or once the concentration values of the nodes are obtained, they can be used to create tropospheric ozone pollution maps with heterogeneous sources (reference stations and low-cost sensors)., Peer reviewed

Ozone concentrations in ambient air collected using low-cost sensor technologies, in the framework of EU project CAPTOR. Data collected during summer 2018 in NE Spain and N Italy. Sensors are metal-oxide. Data are calibrated using multiple linear regression, and validated against official reference data from each local air quality monitoring network. More details on the calibration and data validation may be found in A. Ripoll et al. / Science of the Total Environment 651 (2019) 1166–1179., European Commission:
CAPTOR - Collective Awareness Platform for Tropospheric Ozone Pollution (688110), Peer reviewed

Ozone concentrations in ambient air collected using low-cost sensor technologies, in the framework of EU project CAPTOR. Data collected during summer 2017 in NE Spain and N Italy. Sensors are metal-oxide. Data are calibrated using multiple linear regression, and validated against official reference data from each local air quality monitoring network. More details on the calibration and data validation may be found in A. Ripoll et al. / Science of the Total Environment 651 (2019) 1166–1179., European Commission:
CAPTOR - Collective Awareness Platform for Tropospheric Ozone Pollution (688110), Peer reviewed