Air quality equals quality of life. The challenge: better air-quality mapping. IPI research featured in imec report.

Valorizing the results of multi-year research into air quality monitoring: Flemish technology research on fine-grained air-quality monitoring available to be embraced by government, citizens and industry

Read the full article and report here.

Don’t be mistaken: air quality and quality of life are tightly intertwined than we sometimes realize. Air pollution is considered the environmental factor with the greatest risk for our health and that of our living environment. It’s an invisible threat that concerns all of us. Air pollution is estimated to cause over four million deaths a year because of strokes, heart conditions, lung cancer and chronic lung or respiratory diseases. In comparison: traffic is estimated to result in ‘only’ 1,35 million casualties

The aim and current state of technology research is to develop affordable solutions that can map air quality in a more fine-grained and realtime manner. Basically, one would like to know the air quality on street (or even street-side) level at each moment of the day.

Thanks to several City of Things (CoT) research projects, we have gained the knowledge on what is needed to map these parameters in sufficient detail. Looking at the hardware side, extensive experiments have shown that it is possible to use relatively cheap wireless sensors and still obtain useable and reliable data.

In terms of software, the big-data algorithms introduced higher-up are essential. They need to make accurate and “pseudo-realtime” predictions about air quality at street level from a combination of statically and dynamically collected data. To arrive at this point, imec-nl, together with research teams from UGhent (IPI) and VUB (ETRO), developed and validated machine-learning algorithms capable of interpolating sensor measurements in space and time.

From the moment we can merge all of this knowledge and technology with the expertise and models that are already in place, it becomes possible to determine how many sensors are actually needed and which software should be used to extract relevant information from them.

At this stage of the technological developments, an interesting window of opportunity arises to move further in an orchestrated effort between all stakeholders involved: civilians, governments, research and industry. In an ideal scenario, a model emerges in which civilians roll out and maintain the sensor units and are being supported by government, research and industry to do so and to turn their contribution in to impactful policies and actions.

One of the results of this project is a report title 'Measuring and modelling air quality in smart cities'. This report depicts the learnings from five years of research into the measurement and modelling of air quality in smart cities:

  • With for example the added value and challenges of sensor measurements, the differences compared to reference measurements, sensitivities and applications, insights from data analytics, calibration and validation methods for sensor applications.
  • We offer it as a source of inspiration and a starting point for all stakeholders (civilians, government, research and industry) in the domain to further define and realize the smart city in a collective and iterative way.

This multi-year research was a collaboration between Flanders Environmental Agency (VMM), VITO, IMEC vzw, IPI (UGent), ETRO (VUB), IDLab (UAntwerp) and IMEC-NL and supported by the Flanders Government and VLAIO.

Image Processing and Interpretation