Community air quality monitoring networks
Residents host low-cost air sensors, share the data openly and use the local evidence to push for cleaner air.
Start small
Host 5 to 10 sensors at homes, a school and a community centre along one suspected hotspot, with one sensor co-located next to the nearest official monitor for calibration. Time: 1 month to set up, 3 to 6 months of data. Cost: roughly 300 to 3,000 in hardware depending on sensor (a Sensor.Community kit is around the low end), plus volunteer time. People: one community organiser, one data-savvy volunteer or university partner, 5 to 10 host households, and one named official you want the findings to reach.
The problem
Official monitors are few and far apart, so the pollution people breathe next to truck routes, industry and schools stays invisible and unregulated, and environmental harm goes unchallenged.
Where it fits
Thrives where a trusted community organisation already exists and residents feel a pollution burden officials do not see (freight corridors, border regions, fast-growing cities with little reference monitoring). Needs at least one reference-grade monitor nearby for co-location and calibration, a data platform, a scientific partner, and a policy door to push on (for example California's AB 617 community air plans, city clean air plans, national standards).
The principle underneath
Evidence gathered by the people affected, at the places that matter to them, shifts who gets to define the problem. Dense, open, local data makes hidden harm visible and gives communities standing at the table, as long as it is calibrated well enough to be trusted and tied to a specific decision.
A pattern in nature
Like a lichen survey: many small, sensitive organisms spread across a landscape reveal air quality patterns no single instrument could.
Who to call
West Oakland Environmental Indicators Project (WOEIP); Comite Civico del Valle (Imperial County, CA); AirQo at Makerere University (Kampala); Sensor.Community (originated as luftdaten.info, OK Lab Stuttgart); Breathe London / Imperial College London Environmental Research Group; Environmental Defense Fund air quality team; South Coast AQMD AQ-SPEC sensor evaluations.
Watch out for
Skipping co-location calibration. Picking sites for convenience, not exposure. No plan for maintenance, power and connectivity. Publishing scary raw numbers without context. Having no named decision or decision-maker in mind. Academic partners owning the data instead of the community.
Where it comes from
Grows from the environmental justice and community science movements (e.g. bucket brigades, street-science) and the maker and open data scene. Key holders: WOEIP and EDF in Oakland, Comite Civico del Valle with university partners, OK Lab Stuttgart for the open hardware model, AirQo for African cities. Open practice; respect community data sovereignty agreements where groups set them.
Where it's worked
West Oakland, California, USA · 2015 to 2019
WOEIP with EDF, Aclima, Google and UC Berkeley mapped black carbon and NOx block by block (up to 5x variation on one block; 100 stationary monitors in 2017). The data shaped the West Oakland Community Air Action Plan under AB 617 and truck route decisions.
DocumentedSource
Imperial County, California, USA (US to Mexico border) · Network built mid 2010s; paper 2018
Comite Civico del Valle and university partners co-designed a community particulate network, with residents choosing monitor sites and data sharing, producing data residents trust and researchers can use.
Peer reviewedSource
Kampala, Uganda · 2019 to 2020
AirQo grew from 15 to 40 monitors in greater Kampala, used machine learning to correct low-cost sensor bias, fed into Uganda's draft national air quality standards and helped form the Kampala Air Quality Working Group.
DocumentedSource
London, UK · 2018 to 2025
Breathe London grew from a 100-sensor pilot to over 450 locations, with 60 sensors given to parents' groups, residents' associations and campaign groups; its data was used to evaluate School Streets.
DocumentedSource
Stuttgart, Germany (luftdaten.info / Sensor.Community) · By 2019
Citizen-run open network had more than 300 low-cost PM sensors in Stuttgart and surrounding districts, prompting academic assessment of their reliability.
Peer reviewedSource
Global review (131 studies) · 1990 to 2023, published 2025
About 90% of community low-cost sensor studies focused on sensor performance and only about 10% aimed at policy change, showing the data-to-action gap.
Peer reviewedSource
Where it failed
Low-cost sensors drift and misread, especially PM sensors in high humidity, so uncalibrated networks produce numbers regulators dismiss. A 2025 review of 131 studies found about 90% focused on testing sensors and only about 10% even aimed to bridge research and policy, so most networks generate data that is never acted on. Outside partners who parachute in and leave break trust (a lesson flagged in West Oakland), and projects funded by short grants go dark when sensors fail and no one maintains them.
Related Fruit
- TreeCitizens' assembly on climateA randomly selected, demographically representative group of residents learns from experts, deliberates and recommends climate policy to decision makers.
- TreeCitizen science waterway monitoringTrained local volunteers regularly test their own creek or river and feed standardised data to agencies, catching pollution early and building a long-term record.
- TreeParticipatory community mapping for land rightsCommunities map their own customary territory with sketch maps, GPS and 3D models, turning unwritten tenure into evidence that governments and courts recognise.