Jump to content

Low-cost sensors for monitoring air quality

The monitoring stations measuring the air in Norwegian cities are subject to rigorous quality control, and generate data of high quality.

In recent years, a number of low-cost sensors for measuring air quality have come on the market – so-called micro-sensors. This kind of sensor makes it possible to build dense networks, and when combined with data from the monitoring stations and model calculations, will eventually offer real-time air quality data with very high spatial resolution. However, most low-cost sensors do not yet provide sufficiently high-quality data and still need improvement.

Since 2011, NILU has done extensive research on the use of low-cost sensors and systems for air quality monitoring. One example is the NILU-coordinated CITI-SENSE project (EU FP7 project no. 308524), which deployed what was at the time the largest network of low-cost sensors in the world. In 2015-2016, more than 330 low-cost sensors were simultaneously measuring the air quality in nine European cities, over a period of several months.

NILU’s acquired expertise and experience enable us to provide research-based support to Norwegian municipalities that want to develop low-cost air quality sensor networks and related information solutions.

Low-cost sensor evaluation and validation

Accurate evaluation and validation of measurements done using low-cost sensors is important. Studies have shown that the measurements from today’s low-cost sensors are of varying quality. The sensors, which are intended to measure air pollution in the form of gas and particles, can for example be affected by various weather conditions (wind speed, temperature, humidity) or be “cross-sensitive” and thus unable to distinguish accurately between several different pollutants.

Lavkostsensorer montert på målestasjon i Kirkeveien
Evaluation of the performance of low-cost sensor systems. Sensors mounted at Kirkeveien monitoring station in Oslo. Photo: Jøran Solnes Skaar

NILU has the necessary infrastructure to evaluate low-cost sensor performance against reference instrumentation in both controlled laboratory conditions and real-world field conditions.

Processing data from low-cost sensors

A dense enough network of low-cost sensors also requires the ability to import, store, export and process large amounts of heterogeneous data in real time. NILU is currently building an open data infrastructure to provide scalable “Data as a Service”.

Both commercial and non-commercial actors will have the possibility to make use of a data infrastructure fully designed to make the most out of low-cost sensors.

Real-time high-resolution information

NILU is developing methods to significantly improve the data quality from the low-cost sensor systems. The output can be used to assess local air quality with high accuracy, high spatial resolution and in near real-time, using machine learning and data assimilation techniques. This will revolutionise both the amount and the type of air quality information we can provide to the public.