Pick two
There’s a well-known project management / product management adage… you have three variables – speed / price / quality – and must choose two.
FOSDEM 2025
My friend Chris Adams is, like me1, headed to FOSDEM 2025 – in fact, for the last couple of years, that is the only time I’ve had the joy of seeing him, so I already treasure the opportunity to go there!
Background factoid: Chris and I got to know each other back in the London tech scene, and bonded over topics like the internet of things, energy monitoring and green initiatives – he’s now heading up the Green Web Foundation, as well as co-hosting a very good podcast. He’s also one of the nicest humans I know – and I’m privileged to know a lot of good people.
He’s proposed setting up some very basic air quality measurement in the devrooms – essentially crowdsourcing some data about the concentration in the conference rooms. In his post about the idea, he’s suggested an initial baseline: grab an affordable device from the onlines2, and (with the approval of the room owner) stick it somewhere prominent. Future steps might be to have a nice Grafana dashboard with open data feeding into it.
Suitably nerd-sniped, I have accepted the challenge. This is where that choice starts to bite, though…
Chris is, very sensibly, working upwards from a baseline of “something is better than nothing”. My brain decided to join in, but from the Utopian ideal that we have open hardware and software doing all the work for us already, and that’s where I’ve spent most of Monday morning. First of all – look, yes, I’m as guilty of grabbing cheap stuff from AliExpress as many others, but if I’m going to do so, I’d prefer to know that I can hack it myself and push data to my own MQTT broker.
So, I started by looking into the open hardware options.
There are such devices available. Chris learned about Devtank, a UK-based company that makes some environmental monitoring devices that support MQTT, who I know about via OpenUK events and community. I dug into these a couple of years ago, and my memory is that they are not consumer-level affordable, and may not be the right choice here, but I’m open to being corrected! They also do particulate monitoring rather than CO2, I think.
There are a handful of other full solutions that I know of – the Luftdaten sensors / community, the Apollo Automation Air 1, and the Air Gradient, all of which could be good for this, but are pricier, and more-or-less headless / have minimal displays, thus needing more work to build the dashboard up-front.
There’s the possibility of full build-it-yourself, which could work, assuming folks have the time and energy. The IKEA Vindrikning devices are reasonably priced, and eminently hackable with a Pi Pico W, but again do particulate monitoring rather than CO2. I already have the Pimoroni Enviro+ for both Pico and Pi, once again providing optional PM measurement rather than CO2. That CO2 element is the kicker in most of these cases – it’s solvable with the addition of the correct sensor add-on, but not super quick to do, or as afforable as the cheapo quick purchase options. These would all also want some work to get a nice display in place that was visible in the room, or the provision of a dashboard. I’m in a decent spot – I’ve been collecting and tinkering with these kinds of solutions for years – but this is not a straightforward solution for most of the devrooms.
So… I find myself back at AliExpress, checking on gadgets, and then searching to see whether any FLOSS integrations exist that work – i.e. something with a screen, affordable, supporting CO2 measurement, and a local connection without depending on a vendor’s cloud.
I think some of the things I’m seeing could be persuaded to share their data over BLE to a nearby Pi or other device, and then push that over MQTT, but I’m not so sure that many of them would allow a full firmware replacement that also retains a usable display. I also see some devices that use a Tuya module to do their networking, and there are various states of jailbreaking or sniffing of these online and a Python module, but the level to which they can be reliably detached from an overlord cloud is variable.
All of this is an info dump rather than a solution. It would be great to keep this conversation active, and see where we can get to in the next 3-4 weeks ahead of the event!
I’m likely to try grabbing a couple of cheap (! … ok maybe sub-£30) monitors and see what I can do with them when they arrive, but I don’t really want to put those kinds of things onto FOSDEM network infrastructure without knowing more about how they work.
Right now, I’m stuck at picking two – quick (to procure), low cost; but, proprietary and tied to a dubious data sink – over trying for a better longer-term option.
More ideas welcome! There’s a thread over here on Mastodon, and here on Bluesky; or comment here, if you wish!
I’ll be helping with the Social Web Devroom. ↩︎Chris (and I) do not specifically endorse this one, just an option that he already has.
↩︎ Like it? Share it -
https://andypiper.co.uk/2024/12/30/finding-affordable-air-quality-sensors-for-fosdem/
#100DaysToOffload #airQuality #AQM #carbon #carbonDioxide #ChrisAdams #CO2 #devices #environment #Fosdem #Gadgets #green #GreenWebFoundation #hardware #monitoring #MQTT #Technology
🤘 The Metal Dog 🤘
Andy Piper
🤘 The Metal Dog 🤘
Edward Curtis
Global News
