Making sense of sensor data in practical wireless sensor network designs

Abstract

Fine-grained street-level Carbon Monoxide data with stringent accuracy requirements are required in a collaborative project with environmental scientists. In this project, the chosen electrochemical sensors and the sensor modules were calibrated against concentration and temperature, which is the major environmental factor that affects the baseline of the calibration curve. It was found that there are other environmental factors, including humidity, air flow rate and other gases particular to this application that may contribute to the inaccuracy in the measurements and should be addressed in the system design. The sensitivity of the sensors also decreases with time, which means that sensors are required to be re-calibrated during a long deployment. Observations from real experiments are reported in this paper.