Indoor air quality monitoring & health

So I was looking into a scale which could also measure body fat and I noticed that among its other features, the $150 Withings scale claimed to measure air temperature & CO2 levels. This reminded me of the research on carbon dioxide & cognition (e.g. “Is CO2 an Indoor Pollutant? Direct Effects of Low to Moderate CO2 Concentrations on Human Decision-Making Performance”) among other issues related to air quality. Air temperature seems like it may affect sleep quality, and I recently got a humidifier because I was having problems with eyes burning at night and waking up with a stuffed nose, problems which I recall getting worse than winter when the air around here really dries up. So air for several reasons seems worth measuring.

I became curious about whether there were any cheaper ways than the Withings to measure air quality - $150 is a lot of money especially when there are a number of negative reviews suggesting the Withings scale is not very reliable. (For example, I could instead buy the Omron Body Composition Monitor at half the price.)

I began looking through QS.com, the QS forums, Reddit, Amazon, and Google for devices. It turns out that most discussion centers around either grander topics like asthma or mortality rates or improving personal air quality by buying filters or plants. Looking for measuring devices in particular, it is quite difficult to find a sensor which is not >$150, has a decent selection of sensors, and records data for later analysis. For example, if you look on Amazon for queries like “indoor air quality sensor”, there’s hardly anything under $200 and most of them measure only a few things and you would have to manually record data, and if you google, many product pages don’t even list a price (ie ‘you can’t afford me’). There are a few hyped reports about using smartphone cameras to measure pollution levels, but it’s unclear if any ever made it to a usable app.

The list:

I’m not sure which of these is best. None really stand out. The Kongin is cheap and suitable for tracking air during sleep, but omits CO2. The Withings doesn’t sample frequently. The AirPi has one of the better suites of sensors fully-loaded but it also omits CO2 and putting it together may be very challenging for a novice like myself. Cubesensors and Air Quality Egg are very expensive. Netatmo isn’t that bad looking, but costs too much to be the best candidate; if they came down another $50 or so, they might become the obvious candidate - not the best suite of sensors but at least they do CO2 and it’s easy to use. Smart Citizen measures too little. I’m not sure what the state of the art is for Arduinos since there’s so many pages.

4 Likes

I have been using a Netatmo for 1.5y, works great. Data is logged at 5min intervals, and is available through their API. The Netatmo doesn’t measure VOCs, just CO2. The CO2 sensor calibrates itself by assuming that the concentration drops to 400ppm at least once a week, so make sure you open the window at least once a week :slight_smile: You get an additional “outdoor” station that records just humidity and temperature; this data could be useful, but I haven’t set it up so far because it’s not water resistant… The only thing missing is a light sensor, but I might get a separate (wearable) device for that anyway.

Edited, thanks.

Set it up in your bedroom so you can correlate humidity & temperature against sleep? (I assume you keep the main base in your office or wherever you spend most of your time.)

No, I keep the main station in the bedroom, though trying to correlate air quality with some measure of productivity could be interesting, too.

I actually have the outdoor station in the bathroom now, but am not doing anything with the data (track how much the shower is used?). The home office is out of range of the main station (the outdoor station uses BLE).

Smartphones may be more promising than I thought. It seems that barometers are now near-universal (to improve GPS accuracy), so you can get that off your phone. The Weather Station app says that smartphones are starting to ship with temperature sensors (intended for the battery), and some even have a hygrometer, so that means with some you can get air pressure, heat*, and humidity, which is a good start toward monitoring air quality. It lists as having all 3 the Samsung Galaxy S4 and Samsung Galaxy Note 3; checking on Amazon, the former costs $330, and latter $400-500.

The downside is that this costs more than some of the better options like the Netatmo; but of course, you’re getting a smartphone and not just a temp/humidity logger, the prices will surely come down steeply over time since the smartphone market is so hot, and you have much more control over the device & data than you do for any of the alternatives except the Airduino (and arguably more than it since you get a full Linux OS with pretty beefy specs).

Doesn’t include any of the other sensors like CO2, CO, particulate matter etc, but perhaps there will be plug-in or Bluetooth-based sensors? Some googling suggests the ‘Sensordrone’ project ( https://www.kickstarter.com/projects/453951341/sensordrone-the-6th-sense-of-your-smartphoneand-be / http://sensorcon.com/ambient-co2-sensor-carbon-dioxide-sensor-module-for-sensordrone ) has already done this, has anyone used their devices?

* I realize that a smartphone may not seem to be a good fit since you keep it in your pocket; but air pressure wouldn’t be affected by this, I don’t think humidity would be much affected, and for temperature readings, I would expect it to be not too difficult to statistically estimate in/out-of-pocket state based on temperature/orientation/acceleration and only retain out-of-pocket temperature readings.

1 Like

Looks like http://aircasting.org/ is expanding into a Kickstarter: http://www.wired.com/2014/11/goofy-wearable-tracks-air-quality-around/

There’s no mention of recording any data besides PM2.5 in the Wired article, the Kickstarter page, or the Kickstarter comments. Wearable is nice, especially for people who travel around a lot or live in a city, but $200 is still a lot to pay for a single metric…

The AirBeam could be good at home, although at $200 is placed in the premium end of the market. Would anyone really wear this though?!?!?

If you look at their KS page, it seems at least 110 people do believe they will wear it.

http://www.wired.com/2014/11/clarity-wearable/ “This Wearable Detects Pollution to Build Air Quality Maps in Real Time”:

This sounds far more promising than Airbeam: 6 metrics, not 1; and a quarter to half the price? Sign me up!

Here’s another wearable environmental monitor project on Kickstarter: http://www.mytzoa.com/.

This device has sensors for UV, light, humidity, temperature and air quality. The only drawback is that air quality is measured using PM2.5/PM10, which may not make much sense for indoor measurements. Also, doesn’t look like this device will ship anytime soon.

True, but it can also be used in the home as a standalone device too. It is hard to tell the intended use from this number alone.

Regarding the AirPi, I’ve had similar frustration after building my own - while it works great, my biggest issue is the lack of sensor calibration and conversion of readings to something human-readable. For example, for carbon monoxide or nitrogen dioxide readings are shown in ohms (raw sensor value) and no one has converted these readings to PPM. Likewise, noise level is simply mV instead of dB (and to your original point, it does look like people have added CO2 sensors to their AirPies).

You’ll see that the temperature reading is about 15 degrees F higher than it should be - this is a design flaw with the AirPi. The sensor is picking up heat from the circuit board when soldered directly to it so needed to be separated by about 6" of wire.

Something more “out of the box” that looks promising is Foobot (formerly “Alima” and before that “AirboxLab”) - http://foobot.io/ . I inquired about API access and was told there will be one available when the units start shipping early next year.

1 Like

Nice! I could never get mine to work :frowning:

hi I am using AdvancedSense IQ 610 to analyse my indoor air quality . I always monitor room temperature , humidity level inside the room. I am using this from past 1 year and is working fine for me.

NYT article which mentions a few of the air quality sensors: http://www.nytimes.com/2015/04/16/business/experimenting-at-home-with-air-quality-monitors.html

Products:

Projects:

Awesome thread to find! I’ve been keeping an eye on AirPi, but I’m concerned about the soldering requirements, bugs, and lack of calibration.

I did a bunch of research today on alternatives, and it looks like Netatmo is the current best, at $145 for CSV export and the critical sensors (temperature, humidity, CO2) plus pressure, sound, and IFTTT support.

Foobot would be second, since it has more sensors (temperature, humidity, CO2, CO, tVOC, PM2.5) for $199… Unfortunately, it doesn’t have any export, nor an API, nor IFTTT support.

Is there anything on the market that’s better than the Netatmo, and still under $200? (Everything else from @gwern’s post on 2014/9/29 is either unavailable or more expensive.)

Actually, looks like Foobot now provides API access - http://foobot.io/support/ (scroll down and click “API for Developers”).

I’ve just written a post about my recent indoor air quality measurement experiments (focusing on my bedroom):

http://www.quantifiedbob.com/2015/11/understanding-my-indoor-environment-part-1-air-quality/

As I mentioned in a prior reply, I began by using AirPi, but then switched to Foobot (formerly Alima), which has an API so I can create graphs like these:

Unfortunately, neither device provides all of the readings I am looking for, so in the near term I will need to pull in/combine data from both devices.

Anyone else want to weigh in on Foobot, especially comparisons of accuracy with other sensors? As far as $/sensors go, it seems to be the best available, but the Amazon reviews are extremely polarized between love and hate, with the haters citing both bugs in the apps (not too relevant to us since we’d be using the API, I suspect) but also claiming inaccuracy (a big concern).