Earth Notes: A Note On the Air Quality Egg (AQE): Review

Low-cost Internet-connected air-quality monitor.
old AQE boxes

Review

  • Air Quality Egg
  • Reviewed by: Damon Hart-Davis on 2013/05/24
  • Somewhat flaky, but seems to basically work
  • Regularly samples various air-quality measures such as temperature, humidity, NO2, CO and particulates/dust, and reports those measures live/publicly for access over the Internet.
  • Rating: 3 out of 5

Motivation

I'm keen on biomass as a low-carbon fuel (through its lifecycle) and as a means of storing energy through the year for example to help the UK through cold still winter days and nights when solar and wind renewables cannot contribute much.

I'm less keen on biomass combustion with inadequate flue filtering, reducing air quality and general health and starting to undo the good work of the Clear Air Acts. I'm particularly unkeen on biomass burning in built-up areas (it's possible to detect the combustion products of trendy real fires in Islingston fireplaces on cold weekends I'm told) where the downsides are amplified.

Advertisements

To that end I'm interested in measuring (live) and getting into the public domain and making widely available air-quality data for London and beyond. Thus this mini project!

Initial Findings

Ordered from Wicked Devices in the US 2013/04/30 two AQEs with dust sensor add-on for USD486 total (AQE@USD185, dust sensor @USD58); no P&P charge. Arrived 2013/05/09. First AQE set up indoors and put on-line 2013/05/13 (also visible here). Second AQE set up outside and a little distance away and put on-line 2013/06/03 (also visible here). Decent robust packaging survived international transit!

Immediate bugbears first:

  • The AQE that I have set up regularly loses stops publishing to the Internet (after as little as a few minutes to a few hours) and the sensor and base unit have to be power-cycled together to reconnect. As of 2013/05/24 the system has managed its longest run without needing hand-holding of well over 24h. Could be a problem if a fit-and-forget system is needed.
  • What's with the blindingly-bright lights? My router (into which the base unit has to be plugged) is in the master bedroom and I've had to hide the base unit down the back of furniture and swathe it in dark cloth to avoid it keeping us awake; possibly even a fire hazard. The ability to turn off the light show at least at night would be welcome.
  • Even the sensor unit has lights, but less bright. If I hang the sensor unit outside I fear that it may become a magnet for the light-fingered; no value to them but expensive to replace.
  • Needing to have a mains supply to the sensor unit is a pain; I've just been building Arduino compatible (ATmega328P-based) hardware that runs a variety of sensors and a radio for a few tens of microamps; good to run off a couple of NiMH rechargables for a couple of years. The more complex AQE sensors may make that difficult, but swapping batteries occasionally could well be less pain than the mains tether (and running out of a house with air-tightness measures).
  • The first AQE was showing on their map initially, but as of 2013/05/24 is not, which is worrying. And then a little later as I write, it's back...
  • Inside, lying on its side as suggested, after a while the AQE reads ~5°C over actual temperatures from self-heating (which is also worrying from a power-consumption point of view); this also make the (relative) humidity reading somewhat suspect (low).
  • It does not seem possible to change location of indoor/outdoor metadata after initial set-up, at least through the AQE home-page or similar, which means that redeploying a AQE may mislead users of its public data.
  • The provision of US adaptors (albeit with a 100V--240V voltage range) required me to go and buy a bunch of shaver adaptors.
  • The power-adaptor cables are perhaps a bit short for realistic deployment.

First Steps: end 2013/05

Having observed the AQE running as-is for over a week, my next task is over the next few days at the end of May to try to automate fetching of the published AQE data from Xively and republishing on a local page, live, possibly integrated with additional data from elsewhere.

Measured Power Consumption

  • Base unit: ~2.2W
  • Sensor unit: ~2.8W

Finished

I tried running the AQEs at school to measure temperature and particulates but a number of problems intervened from huge leaf/dander shed from a nearby tree to reconfiguration of the school's networking preventing data egress, and difficulty of retrieving the data via Xively.

As a result the AQE experiment lapsed and the AQEs lie unused gathering dust!

Advertisements