Dataset Information

The Relocation Problem of Field Calibrated Low-Cost Sensor Systems in Air Quality Monitoring: A Sampling Bias.

ABSTRACT: This publication revises the deteriorated performance of field calibrated low-cost sensor systems after spatial and temporal relocation, which is often reported for air quality monitoring devices that use machine learning models as part of their software to compensate for cross-sensitivities or interferences with environmental parameters. The cause of this relocation problem and its relationship to the chosen algorithm is elucidated using published experimental data in combination with techniques from data science. Thus, the origin is traced back to insufficient sampling of data that is used for calibration followed by the incorporation of bias into models. Biases often stem from non-representative data and are a common problem in machine learning, and more generally in artificial intelligence, and as such a rising concern. Finally, bias is believed to be partly reducible in this specific application by using balanced data sets generated in well-controlled laboratory experiments, although not trivial due to the need for infrastructure and professional competence.

SUBMITTER: Tancev G

PROVIDER: S-EPMC7662848 | biostudies-literature | 2020 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

The Relocation Problem of Field Calibrated Low-Cost Sensor Systems in Air Quality Monitoring: A Sampling Bias.

Tancev Georgi G Pascale Céline C

Sensors (Basel, Switzerland) 20201030 21

This publication revises the deteriorated performance of field calibrated low-cost sensor systems after spatial and temporal relocation, which is often reported for air quality monitoring devices that use machine learning models as part of their software to compensate for cross-sensitivities or interferences with environmental parameters. The cause of this relocation problem and its relationship to the chosen algorithm is elucidated using published experimental data in combination with technique ...[more]

PMID: 33143233

Similar Datasets

Project description:Analyses of the relationships between climate, air substances and health usually concentrate on urban environments because of increased urban temperatures, high levels of air pollution and the exposure of a large number of people compared to rural environments. Ongoing urbanization, demographic ageing and climate change lead to an increased vulnerability with respect to climate-related extremes and air pollution. However, systematic analyses of the specific local-scale characteristics of health-relevant atmospheric conditions and compositions in urban environments are still scarce because of the lack of high-resolution monitoring networks. In recent years, low-cost sensors (LCS) became available, which potentially provide the opportunity to monitor atmospheric conditions with a high spatial resolution and which allow monitoring directly at vulnerable people. In this study, we present the atmospheric exposure low-cost monitoring (AELCM) system for several air substances like ozone, nitrogen dioxide, carbon monoxide and particulate matter, as well as meteorological variables developed by our research group. The measurement equipment is calibrated using multiple linear regression and extensively tested based on a field evaluation approach at an urban background site using the high-quality measurement unit, the atmospheric exposure monitoring station (AEMS) for meteorology and air substances, of our research group. The field evaluation took place over a time span of 4 to 8 months. The electrochemical ozone sensors (SPEC DGS-O3: R2: 0.71-0.95, RMSE: 3.31-7.79 ppb) and particulate matter sensors (SPS30 PM1/PM2.5: R2: 0.96-0.97/0.90-0.94, RMSE: 0.77-1.07 µg/m3/1.27-1.96 µg/m3) showed the best performances at the urban background site, while the other sensors underperformed tremendously (SPEC DGS-NO2, SPEC DGS-CO, MQ131, MiCS-2714 and MiCS-4514). The results of our study show that meaningful local-scale measurements are possible with the former sensors deployed in an AELCM unit.

Dataset Information

The Relocation Problem of Field Calibrated Low-Cost Sensor Systems in Air Quality Monitoring: A Sampling Bias.

Publications

The Relocation Problem of Field Calibrated Low-Cost Sensor Systems in Air Quality Monitoring: A Sampling Bias.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets