Affordable sensing devices can widen and deepen air quality monitoring by targeting areas not covered by official measurements. However, low-cost sensors have been criticised for producing data of questionable quality. The webinar explains how this issue was tackled in COMPAIR using a distant, cloud-based calibration algorithm. Another issue that can affect citizen driven air quality monitoring is network connectivity. Some low-cost sensors depend on specific network standards, such as NB-IoT, for data transfer. During the webinar, COMPAIR shares recommendations for minimising disruption to citizen science in cities where the necessary IoT infrastructure is lacking. Finally, the webinar shows how the integration of diverse data sources, including traffic counts, into citizen science can lead to a more holistic policy impact assessment.
Citizen science for local policy making: opportunities and challenges
Citizen science can mobilise communities to act on urban challenges like climate change and air pollution. Participants gain new knowledge and skills that help them change their behaviour towards more sustainable lifestyles. Citizen science can also support policy making by providing high-resolution data on air pollution in areas not covered by official monitoring stations. However, due to quality concerns, the uptake of citizen science results by policy makers remains low. One way to improve trust in citizen science results is to enhance the quality of collected data. This is where calibration methods come in.
Citizen science sensors: improving data quality with calibration techniques
To comply with environmental regulations, cities use expensive, highly accurate monitoring stations to measure air pollution. But given their sparse distribution, inner-city pockets and hard to reach areas are poorly covered. Citizen science can address these gaps by providing hyperlocal data on air quality. The main issue with citizen science sensors, however, is that they typically use low-cost components that are sensitive to environmental conditions and can therefore over- or underestimate air pollution. Cloud calibration techniques provide a means to correct for these inaccuracies.
Best practice and lessons learned from citizen-led traffic monitoring
In citizen science projects with a significant ICT component, there is a natural tendency to over-engineer technical solutions. However, participants generally don’t want and need complex tools. The higher the technical barrier, the smaller the pool of potential participants, and the higher the likelihood of dropouts. Sensors and apps should therefore be designed with ease and self-installation in mind. In addition, trials and workshops should aim to “debug” not the technology but user journey first and foremost, ensuring that key participants’ expectations regarding process and benefits are met.
Variability of IoT network connectivity and risks to citizen science
In most European countries, network coverage for NB-IoT and LTE-M - two standards necessary for data transmission in some IoT devices - is good. However, in the Balkans, for example, the connectivity is poor, which hinders the rollout of IoT sensor networks that rely on these standards. There are also noticeable differences between cities and the countryside. If the necessary IoT infrastructure is lacking, data collection will not be possible, which will have a knock-on effect on citizen participation. An early risk assessment is needed to understand which devices can and cannot be used in the area of interest to minimise possible future disruptions to a citizen science experiment.