The New Project Using ‘Boris Bikes’ To Map Air Quality


Salt welcomes new columnist Jeremy Green, principal analyst at Internet of Things and Machine to Machine advisory company Machina Research. He is an expert industry commentator, especially in the sustainable transport sector.

collumnistAirPublic, a project which mounts air quality sensors on Transport for London (TFL) hire bikes and uses the data gathered to provide denser, more granular and more informative air quality maps for the city, has been awarded seed money as a result of a hackathon held at Climathon, an event organised by the EU-funded climate innovation network Climate-KIC.

AirPublic proposes to put sensors on London’s rental bikes – and eventually bikes in other urban hire schemes – to fill in the gaps in the official air quality sensor networks.

Everyone knows that the air quality in London, as in many other big cities, isn’t too good. It’s regularly reported how bad London air quality is, and there is increasing awareness of the impact of this on public health and the economy. The air in the UK’s biggest cities breaches EU legal limits, and it’s not going to get better any time soon.

But it’s not generally appreciated that air quality monitoring depends on a small network of sensors. London Air Quality Network, which is used to test for compliance against air quality standards, includes only 120 sites, measuring oxides of nitrogen (NOx), sulphur dioxide (SO2), ozone (O3), carbon monoxide (CO) and particles (PM10, PM2.5). These sites provide high-resolution hourly information which is then posted via web and other platforms. These monitoring stations are run by the Environmental Research Group at King’s College. This has its own app which provides air quality information, and regularly updated maps; but most of what the map shows is derived from models, based on the data from the 120 sites.

The AirPublic project is premised on the argument that this modelling is not all that reliable and may understate just how bad London’s air is. It says that a community monitoring initiative supported by the London Sustainability Exchange showed that the actual pollution levels in some locations were worse than those predicted by the model; the model doesn’t account for the ‘canyon effect’ and other factors that makes air quality in Oxford Street arguably the worst in the whole world, for example.

Mounting sensors on bikes will massively increase the number of sampling points. Even a pilot deployment on 50 hire bikes would cover 2,500 out of London’s 2,500km of main roads every day. Taking a measurement of pollutants every 10 seconds/minutes (around every 20 metres) would thus generate an estimated 1,296,000 data points from different locations every day. The bike sensors would be cheaper, less accurate sensors than the heavy-duty ones in the London Air Quality Network’s stations, but there would be a lot more of them. What’s more, they could calibrate themselves against the more accurate monitors every time one of the bikes pass by one of the King’s College sensor sites, thereby improving the overall quality of the data.

Other exciting ideas include allowing users of individual bikes to track their own pollution exposure in real time and perhaps presenting users of TFL’s bike route planning tool with ‘clean’ and ‘dirty’ options as well as ‘busy’ and ‘quiet’ choices.

AirPublic is far from the only air quality monitoring initiative in town, of course. There is also:

  • AirSensa, a project of ChangeLondon which is intended to be a business-funded network of 10,000 high-spec sensors across London.
  • BuggyAir, which aims to mount sensors on strollers so as to inform parents about the levels of pollution to which their young children are exposed.
  • Pollupa, which overlays data from the London Air Quality Network on the Zoopla property sale map, so that prospective buyers can see how polluted the air is at their potential purchase; Pollupla also won seed money at the same Climathon as AirPublic.
  • A host of citizen science and community projects, summarised here in the blog of the Muki Haklay, the doyen of Citizen Science. Some of these are based on low-cost community-built hardware like the Air Quality Egg.

At the moment AirPublic is mainly an idea, though it’s a well defined one that others have judged worthy of funding. It’s not yet certain what organisational form it’s going to take, though it will have to resolve this in order to receive the money from the Climathon.

Its first deliverable will be a technical proof of concept, to be ready in time for the Sustainable Innovation Forum scheduled to precede the global climate talks in Paris this December. The sensor hardware is not yet defined, let alone built, though the presence in the AirPublic team of Tom Hartley, the wonderkind responsible for the low-cost RaspberryPi based AirPi monitoring device (an alternative to the Air Quality Egg), must be reassuring in this respect. The open source Internet of Things start-up has offered to host the data and provide analytics for free, and there has been some discussion with OpenTRV about using the latter’s Lora-based IoT network to provide connectivity for the sensors.

After the proof of concept demonstration there will probably be a pilot involving some 50 bikes in London, followed by a full rollout with 500 bikes.


Photo credit: Simon Thomas from flickr