Routing around pollution – any help?

While we’re benefiting from all this rain in London, which keeps air pollution at bay, I’ve been wondering about including pollution data in OpenStreetMap-based routing engines. The trouble is that I lack the technical skills to implement this, so I’m writing this post in the hope that somebody might be inspired to give it a go.

Another hazy air pollution episode in London

Air pollution is the second biggest cause of premature deaths in the UK after smoking. Here’s a little league table of nasties taken from Department of Health data:

Smoking – 87,000 premature deaths per year
Air pollution – 29,000
Alcohol – 22,000
Obesity – 9,000

The main pollutants in cities are particulates (PM10, PM2.5), nitrogen dioxide (NO2) and nitrogen oxide (NO). Unlike the pea souper fogs of years gone by, these are invisible to the human eye but very deadly. They’re most concentrated on busy congested roads and around airports – so unsurprisingly in central London, around Heathrow and City airports, and in congestion hot spots dotted around the rest of London. We have some of the worst air pollution in Europe.

Long term exposure to these critters is obviously a bad thing, so it would be nice if we could find routes to walk or cycle without being affected. To some extent it’s just a matter of avoiding main roads, but some are much worse than others and in central London there are plenty of non-A roads that should be avoided as well.

WalkIt has a neat little feature that lets you choose a “low pollution” route. This basically tries to avoid roads with high concentrations of nitrogen dioxide. But it doesn’t cover the whole of the city, it doesn’t take account of nitrogen oxide or particulates, and it’s only for pedestrians.

So… who fancies taking the London Atmospheric Emissions Inventory maps, delivered in 20m grid shapefiles, and plugging them into one of the various OpenStreetMap routing engines to provide a walking & cycling “avoid polluted roads” option for London?

Bonus points could go to anyone who uses live data from the London Air Quality Network and Defra’s air quality monitoring web site (with its latest readings from a few monitoring stations) to give more or less priority depending on conditions.


  1. sanderd17 said:

    As pollution data is only available for certain areas, and you are from London, I think cyclestreets might be interested in this.

    15th May 2012
    • Tom Chance said:

      Thanks, I’ve suggested it to them already, fingers crossed. That would still leave scope for a more general solution that could help pedestrians, motorcyclists, horse riders, etc!

      Also there’s no reason further pollution data for other cities couldn’t be pulled in, I’m sure most European cities with pollution problems will have comparable publicly available data given that it is measured (and hopefully tackled) to comply with European regulations.

      15th May 2012
      • sanderd17 said:

        This will be a huge task, as each city will provide the data in their own format, with their own licensing terms etc.

        Unless you get help from a local government, this won’t be do-able.

        15th May 2012
      • Tom Chance said:

        Never say never! The same could be said of all data in OSM, it’s possible if someone were really interested.

        The London data is very simple, just a shapefile with a 20m grid that could be overlayed onto the road network to provide a set of pollution values for segments of roads.

        15th May 2012
      • cyclestreets said:

        Will contact you off-list shortly!

        15th May 2012
  2. Seems like a similar problem to making routing system which avoids current traffic. Someone was pointing out to me that google offers live traffic overlays, but does not offer routing to avoid traffic. I guess it’s a difficult problem because you’d have rebuild your indexes a lot. Or maybe air pollution is less changeable? In which case it’s really just a matter of weighting routing accordingly. Systems do this already based on (fixed) speed limit info.

    15th May 2012
    • Tom Chance said:

      One could either start with a simple, static picture, taking the data I suggested in my original post which is for average levels through the year. That’s obviously what WalkIt have done. This would basically do the trick, any interface to the routing engine could alert users to the point that pollution isn’t a big deal if it’s raining, but it’s a huge deal if you’re in the middle of a long period of high pressure, no wind, warm weather.

      Or one could get clever and take live updates, either for warnings (like traffic) or to actually re-index.

      15th May 2012
  3. Sven Geggus said: has an avoid area feature.

    15th May 2012
    • Tom Chance said:

      Interesting, thanks.

      15th May 2012
  4. sanderd17 said:

    I just went on quest for some Flemish data. And this is what I found:

    Although the data is (legally) usable, there are far too few measuring points to be useful. Maybe only Antwerp could get some routing like you suggest.

    15th May 2012
  5. […] direct walking and cycling routes that keep you off the most polluted roads. I blogged about this last year, and I still think it would be both cool and genuinely […]

    12th September 2013

Comments are closed.