fiddyspence's blog


If you’re reading this, you’re at least half interested in community based monitoring of airport noise. The end result of this should be noise heatmaps around (initially) Heathrow that provide time based indication of peak/average noise around the airport. As an aside, there might be some useful data produced and there might also be some reusable software that can be iterated on for residents around other airports.

In the initial phase the stack to actually monitor noise consists of a few cheap bits of hardware plus some software. There is no particular expectation that anyone needs to be able to write computer programs, though given the initial lack of real packaging around the stack, a bit of Linux sysadminning and familiarity with computering will come in handy.

The stack consists of a Raspberry PI (though any small computer will do, providing it’s capable of running Ruby) plus a USB attached sound pressure meter. I’ve been using a Wensn 1361, though any other USB attached similar device should do. In total, the hardware costs about 75GBP and requires a little bit of time to configure it. Assuming this proof of concept works, then I’ll put some effort in to packaging a proper distribution with some tooling to make it easier (and depending on demand bulk acquire the kits and save a bit of money that way). It’s preferable that the kit be as close to outside as possible so as to not mute the noise - I don’t think it’s terribly useful to graph internal sound - it’s really the patterns of peak/trough external noise that will be interesting. Background levels are interesting too. Eventually, assuming success, it would be good to get access to CAA or other flight path data and be able to correlate peak noise with particular times/flights/aircraft/airlines/geese.

The purpose of the effort is to capture time series data, bound to known locations, that can be used to produce metrics and visualisations of noise. Initally, I want to be able to generate heatmap visualisations overlaid onto google map data, with the enough coverage for it to be meaningful, and be able to view different noise profiles over time.

If this sounds interesting, if you have further questions, or something you’d like to get involved in - either running a noise sensor, contributing to the software etc, then contact me through twitter - @tophlammiepie