Environment-Adaptive Calibration System for Outdoor Low-Cost Electrochemical Gas Sensors

Abstract

To date, air pollution has caused over seven million deaths worldwide. Real-time monitoring is an efficient solution to prevent and mitigate air pollution. However, air quality monitoring networks are not currently dense enough to cover a large area because of the cost. With the development of low-cost electrochemical gas sensors, the current air quality monitoring network could be extended to a large area with a relatively high-resolution ratio. Nevertheless, the electrochemical gas sensors are not accurate enough compared to standard sensors if they are not carefully calibrated. Therefore, we proposed an environment-adaptive calibration (EAC) system dedicated to the implementation of a collaborative calibration technique for outdoor low-cost electrochemical gas sensors. The EAC system uses a new practical calibration algorithm operating on the Raptor IoT cloud platform that was developed as part of the H2020 Captor project. The system can identify the linear characteristics of gas sensors according to weather conditions, which is valuable prior knowledge for the calibration and the unique character of the EAC system. The outstanding performance of the EAC system is verified by the evaluation of a long-term measurement campaign, in which various metrics are considered, e.g., the goodness-of-fit, target diagram, and daily residuals.

Keywords

• Air quality monitoring
• environment-adaptive calibration (EAC)
• IoT cloud platform
• low-cost electrochemical gas sensors
• non-linear regression