Hacker News new | comments | show | ask | jobs | submit login
Show HN: Low cost CO₂ monitoring (your office) with Prometheus and Go (github.com)
127 points by leastangle 8 months ago | hide | past | web | favorite | 35 comments



This is another way to do IoT wrong - collecting data but just displaying it.

The idea is to use CO₂ levels to control your HVAC system. The HVAC system should be able to draw fresh air from outside or recirculate air from inside, depending on CO₂ level. This is a standard option on modern HVAC systems, and there are standard sensors for it. Better systems sense temperature, CO₂, CO, humidity, and smoke. This is a huge win for rooms where the people load varies widely, such as hotel function rooms and classrooms. Such control systems save money, because, when nobody is using the room, they detect that CO₂ is low and cut down the ventilation rate. When the room fills up, the CO₂ level goes up, the fans speed up and the outside air intakes open until the CO₂ level comes down.[2] There are smart control units which manage heat, fans, vents, and air conditioning compressors. The hardware pays for itself in power consumption.

And yes, you can get this stuff Internet-enabled, although that's mostly for remote maintenance. The HVAC works just fine without connectivity.

Surprisingly, this technology is a tough sell. Except for convention hotels. They get this. They're in the competitive business of keeping large numbers of people comfortable and coming back. They have big rooms where the people load may go from zero to a thousand in minutes, and go back down an hour later. It's a huge win for them.

[1] http://www.airtesttechnologies.com/support/reference/CO2SeqO... [2] https://buildingcontrols.honeywell.com/literature/Advanced_R...


> Surprisingly, this technology is a tough sell.

Well, in the Office for regulating CO2-vs-O2 it's competing with the timeless, very-low-maintenance and low-cost combo of keeping-plants-and-occasionally-opening-a-window.


If the windows and the HVAC aren't talking, you end up trying to heat or air condition the outside world. This is both futile and expensive.

Also, it takes about eight trees to absorb the CO2 emitted by one human. The contribution of houseplants is tiny. Especially since they usually don't get full sunlight.


>>> If the windows and the HVAC aren't talking, you end up trying to heat or air condition the outside world. This is both futile and expensive.

That's an argument to be made to the hotel or office owner. The user has no control on the HVAC and is forced to open the window or turn on electric heaters to cancel it.


If you're interested in a hardware solution, I've used the MH-Z19 CO2 sensor to great results. It's an NDIR (read:not electrochemical) sensor with a UART interface and temperature compensation. It will report up to 5000ppm and comes factory calibrated.

It's also $20 in singles from China. Coupled with whatever microcontroller you want, it's totally possible to have a distributed net of CO2 sensors for a low cost per node.

That being said, it's a ton of work and this is an excellent solution with much quicker results.


Just wanted to add that Tasmota[0] (one of the open source firmwares for ESP8266 controllers) seems to already support MH-Z19 sensors.

This cuts down on the work if all you want is to measure your CO2 level: you can get a $3 Wemos and a sensor, (solder pins, figure out the sensor connector), flash Tasmota, and you'll have it already reporting on MQTT over wifi. I would argue it's even easier than installing and maintaining Linux on a RPi.

[0] https://github.com/arendst/Sonoff-Tasmota


+1 for the MH-Z19.

We're looking to add a Co2 sensor to our IoT lineup and the MH-Z19 looks promising. It's also really small too :). You can also choose from 2 different ppm sensitivity from the factory and can get it on Aliexpress.

Here's the project I'm working on:

http://www.kokonaut.com http://www.instagram.com/kokonautweathersensors


Thank you for the tip! I was just thinking of adding a CO2 monitor to a homebrew Arduino weather station. I was looking at this device but calibration of CO2 sensors is an issue:

https://www.adafruit.com/product/3566?gclid=EAIaIQobChMI9uiy...


Is there any significant difference between MH-Z19 and MH-Z19B?


I was looking in to a indoor quality monitor, and seems like for $80, you could just buy this which connects to homekit https://www.amazon.com/Elgato-Eve-Room-technology-Bluetooth/...


Came here to say something similar. I expected an actual circuit build (of which there are many), but this is just plugging in a USB data logging CO2 monitor (which is $116 on Amazon right now, but apparently the normal price is ~$70 USD) and using it with Prometheus. I've done something similar wrapping https://github.com/merbanan/rtl_433 to pick up temp / humidity monitoring data from sensors around my home, then stuffing the results into InfluxDB and using Kapacitor for alerting (sends me an email when my freezer gets too hot) + Chronograf for graphing. It took an evening of reading docs and messing around to get it all working, and I'd recommend Ding I Y if it sounds cool -- you'll very likely learn something. Note that you could do the same with an ELK stack or Sensu or Nagios or whatever you want to play with.


What kind of 433MHz temperature+humidity sensors are you using? What kind of battery life do you get from them in the freezer?


They’re all Acurite sensors with lithium batteries (Energizer ultimate lithium, available at Target on sale sometimes and other places). The one in the freezer is a “00592TXR” and lasts ~9 months in there IIRC (lithium cells do much better with cold temps than alkaline). I also have a few model 06044 which are small temp + humidity sensors with a display. These also have good battery life. Lastly, Acurite makes a smart hub to log to their cloud, model 09150M. I haven’t set it up yet, but it basically does the same stuff: https://www.myacurite.com/#/login

Here’s the kind of stuff it can do: https://www.acurite.com/access-my-acurite-remote-monitoring....

I’ve also had an Acurite display with some segments of digits that looked funky depnending on viewing angle. It was usable but kind of bugged me. I contacted them and they RMAd it quickly and easily for me (they just sent a new unit). For this reason, their easy availability on Amazon, the reliability and battery life I’ve had with half a dozen of their sensors deployed in and outdoors, I’ll keep buying them when I need this type of thing in the future. They don’t ever lock up or need fiddling with, so they’re kind of boring in that they just keep working until the batteries go flat. The widespread open source support is also certainly nice :)


Thanks, that's encouraging! I was worried they might not last that long, especially at low temperatures.

I am using a Sonoff RF Bridge[0] to talk to 433MHz sensors. It is also supported by Tasmota[1].

[0] https://www.itead.cc/sonoff-rf-bridge-433.html [1] https://github.com/arendst/Sonoff-Tasmota/wiki/Sonoff-RF-Bri...


AcuRite Access looked very interesting, until I found that it needs a phyical cable to the router. I am trying to move away from cables, as does most of the world. A bit disappointing for a 'New !' product.


Which does not monitor CO₂ content. VOC mostly is an issue with new furniture and can lead to Sick Building Syndrom. CO₂ is an independent issue you will always have with people in the room that breath.


I'm sure it's just a coincidence, but Amazon just bumped the price of that sensor up by almost $40. Perhaps that is a proxy for how popular this article is and how quickly Amazon notices the trend.


Honest question - Why would one want to monitor CO₂ in their office?


Elevated CO2 levels have a noticeable impact on human cognitive ability. Indoor environments obviously have higher CO2 levels vs outdoor, but certain offices may have worse than typical indoor levels due to poor ventilation.

https://thinkprogress.org/exclusive-elevated-co2-levels-dire...


The linked article is fascinating.

"They found that, on average, a typical participant’s cognitive scores dropped 21 percent with a 400 ppm increase in CO2."

If this result is reproducible, this is quite concerning considering:

"In surveys of elementary school classrooms in California and Texas, average CO2 concentrations were above 1,000 ppm, a substantial proportion exceeded 2,000 ppm, and in 21% of Texas classrooms peak CO2 concentration exceeded 3,000 ppm."


Interesting, I had no idea. Do you think it's worth getting a CO2 monitor for a home office? I don't usually have the windows open since it's cold out. Is one person enough to elevate a small 10x10 office to cognitive impairing levels?


It’s a small investment and probably worth it if you already suspect bad ventilation. Within seconds of plugging a CO2 monitor into my bedroom outlet, the monitor was registering 1200ppm — and that was without any humans in it prior.


I doubt you'll need one. It's typically a concern when sqft/person is < 80, as is often the case in open-office plans and classrooms.

Just try to open the window at least once a day.


I disagree. It _may_ be totally worth it depending on the place. (Eg. what heating system you use, how many electronic devices there are, how big the room is.) I'd say do a little test: open the windows; if (it feels refreshing) { get one }.

Disclosure: I'm not an expert, simply got interested in this lately as I strongly feel its impact on my productivity.


Do the huge corresponding drops in CO2 and temperature at 11pm and 9am mean that the "solution" was to open a window or door for ~15 minutes a couple times a day?

Seems like a simple solution, other than there's a good half-hour of sub-17 degree temperatures right at the start of the workday, which isn't exactly a great motivator to start working. Unless you are skilled at typing with winter gloves on, I guess.


Well $80 dollar is affordable but certainly isn't low cost to me. Why are all Co2 Monitor cost so much?

And on the subject, do we have O2 level monitor as well?


What do you guys think about ozone? I know its banned in California from what I've heard. But I hear it kills a lot of bad compounds like mold. I could be wrong though, would love to hear you guys opinion on it.


In almost every office I was there was a poor ventilation. I have extremely high co2 ppm in my university labs. I try to always open windows where it possible.


> we were now able to optimize HVAC settings in our office (Well, we mainly complained to our facility management).

Time to go shop some plants?


A random google link says that you would need hundreds of plants per person to get sufficient oxygen production [1].

And during night, the plants would just produce CO2, which would presumably get accumulated, making the whole situation worse.

[1] - https://io9.gizmodo.com/5955071/how-many-plants-would-you-ne...


Another random google link says it could be as little as four shoulder-height areca palms: http://www.secrets-of-longevity-in-humans.com/oxygen-produci...


Plants have have been shown to capture other contaminants in the air as well, I haven't seen any capture capacity over any time scale. I wonder if anyone has done the metrics.


Plants did not get approved ;)


i am looking for something similar but for particulate matter instead of co2 like PM2.5. I am also looking for a solution that would work mobile, eg. on a bike. Any help greatly appreciated!


There is a project in Germany which monitors particulate matter. They provide a guide how to build a sensor using a SDS011 and a ESP8266: http://luftdaten.info/en/construction-manual/

Not sure if/how this would work mobile, though.




Applications are open for YC Winter 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: