Re: [crash] Interested in helping out with developing an air quality monitoring system?

63 views
Skip to first unread message

Jeffrey Peacock

unread,
Apr 9, 2012, 5:23:56 PM4/9/12
to safeca...@googlegroups.com

Not withstanding the Egg, I think we need to step back and ask:  what are the metrics an "air quality" sensor would sense?  In a conversation with Levi, radiation came up as phenomenon with contemporary interest.  I suggest that it would be better to speak of, and distinguish, such an envisioned device as a low cost "environmental" sensor.  Perhaps even a tri-corder like device?

Obtuse tangent that I hope to dove-tail into the current topic:

For the last couple of years I have been experimenting with a variety of sensors and collecting and managing the (heterogeneous) data and finding ways to correlate the data with other data.  (The same system that collects weather data also collects and manages the cosmic ray data Levi and I are collecting using Android devices.)  There are some practical issues with respect to the sensors themselves:

Fixed based stations:
There are 2 weather stations in my front yard.  One is built from scratch using 1-wire components (from here: http://www.hobbyboards.com) the other is a Davis Vantage Pro2.  The 1st is relatively cheap, the second relatively expensive.  Each can be embellished with other sensors, e.g., ultra-violet radiation.  Each system has been adapted to communication with a small PC datalogger.  The 1-wire station is connected by a 4-lead cat-5 connection from the PC thru the usb-serial 1-wire adapter into a power injector (necessary for one of the sensors which can't be powered by power off the serial line) and then to the station.  The Davis is "wireless".  The station is powered by a battery recharged by a solar panel.  It is fan aspirated which has a separate battery/solar panel power arrangement.  It reports data via proprietary wireless signal (900MHz?) to a console (battery operated with AC adapter).  From there I applied the following hack (http://www.wxforum.net/index.php?topic=10721.msg104689#msg104689) to enable XBee communication from the console to the PC datalogger.  The PC datalogger is the hardware agent that uploads the data to the cloud.

There's a lot of detail above but the critical things I learned from all this:
1.  The hardware/station needs to be self powered -- i.e., have some battery/solar-recharging mechanism.
     I researched doing this for the 1-wire station but at the time (a couple of years ago) I could not find a simple kit for this.  I don't know if one exists now.

2.  Data reporting needs to be wireless.  Wiring interferes with the hardware placement.
     Stations need to be able to be placed in environments that are free from artifacts that will skew their values.
     Draping a wire out of your window is a pain as it tends to let weather and bugs in.

3.  Calibration:  stations need simple field (re-)calibration/maint.
     3a. If possible, They need to be self aware when they are out of calibration.

Mobile based stations:
I also have a Kestrel 4500 Bluetooth enabled handheld weather station (http://www.nkhome.com/kestrel/kestrel-BLUETOOTH/) weather stations provided a kind of reference device I could use to experiment with fixed and mobile stations).

In this scenario the Kestrel device is battery powered and not self replenishing.  The device has local memory which can be used for datalogging but the data is useless unless it can be offloaded and used in a larger analytical context.  My Android phone acts as the wireless datalogger, which initially buffers the data onto a large SD Card and then, if connected to a cell or wifi network, can upload the data the the cloud based datalogger/larger-context-analyzer.

Critical things I learned:
1.  The scenario described shows the advantages of a mobile solution than can be used to monitor conditions in a "hotspot" or other area of interest.  The user can place/anchor the device in the inclement condition and record data from an offset location -- imagine getting live reading of thunderstorms, tornadoes or hurricanes while sitting safely in your car, storm cellar or bunker; or attaching a station (other environmental?) to the roof of your car and riding around while sampling.  Also, data can be embellished with location data from the phone -- i.e., GPS -- accurate time information from the phone -- tends to be more accurate and fresh that the collecting device -- and other sensors from the device if appropriate -- camera can be considered a sensor with respect to recording visual information; audio can be considered in recording noise volume or even lightening-to-thunderclap times in combination with the camera or light sensor.

2.  The Bluetooth is not an ideal wireless mechanism for small devices in that it's power requirements are too great for extended operation -- this changes the battery and replenishing requirements.  XBee would be better.  However, it is not typically avail. on retail devices or phones, which make great network connectivity devices.  For this reason it is probably not suitable for a mobile sensor unless there are some adapters in the mix.  (I have not yet experimented with the XBee explorer on the Android phone via the USB port but I think it possible.)

3.  Managing "session" based (e.g., non-continuous) data in the context of other data being collected.

So, in the discussion of designing an "air quality" or "environmental" sensor these are some of the things I would consider as being part of the design discussion.

*****

Current research/activities:
Aside from the DECO project and wrangling the diversity of the Android hardware WRT the camera idiosyncrasies, which has become the primary project for sensor data and data collection and management, I did finally discover Arduino and purchased the inventor kit, the geiger counter kit, radioactive material(!), "Building Wireless Sensor Networks" by Faludi, and the companion sensor kits that Sparkfun offers, with the intent to prototype a self contained environmental sensor.  Unfortunately I don't have anything to report on that sub-project yet as my day job keeps interfering.

Oh, did I mention I was interested in participating?

/J


On 04/09/2012 12:26 PM, Sean Bonner wrote:
Yeah, that's the Pachube guys project.

On Mon, Apr 9, 2012 at 12:24 PM, Chris Horton <caho...@gmail.com> wrote:
On a related note, just saw this:

http://www.kickstarter.com/projects/edborden/air-quality-egg

-c


On Wed, Apr 4, 2012 at 1:55 PM, Ariel Levi Simons <levis...@gmail.com> wrote:
Hey all,

Project Safecast (http://www.safecast.org) is now starting to branch out to more environmental metrics.  One of these is creating a platform to sense, record, and map air quality data; hopefully in real-time.  Interesting in helping us all brainstorm how to do this?


Thanks!

--
Ariel Levi Simons





--
Sean Bonner

* email is painful so I'm trying to keep messages short and direct. Please excuse my brevity. Here's 6 tips to make email suck less for everyone.

Ariel Levi Simons

unread,
Apr 10, 2012, 11:01:28 PM4/10/12
to crash...@googlegroups.com, safeca...@googlegroups.com
Peter,


That sounds pretty interesting.  So far the most reliable air quality sensors I've seen so far are from Figaro.  Here's their sensor list by target gas:
I'm interested to see what kind of hardware the Air Quality board of California is using for their monitoring.
What's this self funding method you speak of?  I'm interested to hear more on that.

I'm down for physically meeting over this project.  I'm based out of Koreatown in LA and work in west LA.  CRASH Space in west LA may be a good physical meeting space for this kind of work in the near term.  Any thoughts on that matter?

On Tue, Apr 10, 2012 at 3:45 PM, Eric Fort <eric...@gmail.com> wrote:
Can you list sources for each of the sensors listed below?

On Tue, Apr 10, 2012 at 2:28 PM, pete-crashspace
<pete-cr...@peterbenjamin.com> wrote:
>
> I'm very interested in this project and will contribute.  I'll add a list of
> sensors used by the Air Quality Board of California for freeway fume
> monitoring, and a list of sensors the EPA uses for natural disasters, like
> 911.  The initial list of chemicals to be covered are:
>
> Oxygen
> Carbon Dioxide
> Carbon Monoxide
> Sulfur Oxides SOx
> Nitrogen Oxides NOx
> H2S - sewer gas
> Ozone
>
> where for this I know sensors cost between 20 to 40 dollars for low
> sensitivity, ppm range.  For ppb range, sensors go up in price between 50 to
> 200.  It's not for everyone, but it would be good to have a few of each in
> neighborhood.
>
> There are also the smog measurements, both the initial smog chemicals
> measured by California, which included non visible molecules, which were
> removed from the list measured.  Also, high altitude smog components known
> to be even more toxic are not measured.
>
> Also, jet fuel fumes for near airports.  And for the prevailing downwind
> direction from factory sections of town, we should add toluene, acetone,
> alcohol, and other VOC's, cleaners, known to be in use at those facilities.
>
> Asbestos is a more difficult one, and should be rotated to those stations
> downwind of building being demolished.
>
> So, creating Google map zones, or our own OpenGIS website, listing sources
> (based on government and private industry web published lists) and their
> prevailing "plume" direction and size, for recommending types of sensors
> in/under the plume, as well as displaying real time results.
>
> I know a self funding method after we have several stations in operation.
>
> I'm all for regular monthly meetings, for face to face, design and implement
> and test meetings.
>
> It would would be good to become a certified calibration lab in the awesome
> state of California, so we do not have to go Weights and Measures department
> labs, and pay their fee rates.
>
> BTW, these sensors come with boards that are Arduino compatible!!!
>
> Or the raw sensor can be hooked directly to the Arduino.
>
> Better, is bulk buying likely can cut the price down.
>
> Peter
>



--
Ariel Levi Simons

Jeffrey Peacock

unread,
Apr 11, 2012, 12:22:54 AM4/11/12
to crash...@googlegroups.com, safeca...@googlegroups.com
From the hip:

o  Is it useful to measure O2 in open spaces?  It tends to vary very little.
o  Which sensors are solid state and which use consumables or have very
short life spans?

/J

Ariel Levi Simons

unread,
Apr 11, 2012, 10:01:38 AM4/11/12
to crash...@googlegroups.com, safeca...@googlegroups.com
From my experience it's generally not very useful to measure oxygen levels unless you're measuring concentrations down to the parts per million level.  Since most monitors I've seen only measure down to parts per thousand it is usually pretty pointless to directly measure oxygen in the first place.

All of the sensors I've seen with main manufacturers such as Figaro, and to a lesser degree Futurelec, are robust solid state devices which don't require any sort of consumables.  Although I don't know what the lifetimes are of the devices supplied by Figaro they should not be put near any object with silicone as its vapors will ruin the sensor junctions.
--
Ariel Levi Simons

Ariel Levi Simons

unread,
Apr 14, 2012, 7:46:07 PM4/14/12
to crash...@googlegroups.com, safeca...@googlegroups.com
Hey all,

If you're interested in working on adding air quality metrics to the Safecast platform, so we can create air quality maps along with radiation maps, please come on by CRASH Space this coming Tuesday at 6pm.  We'll be doing our kickoff meeting for Safecast-air.
What will we be talking about and working on?  Please check below:

On Fri, Apr 13, 2012 at 9:23 PM, pete-crashspace <pete-cr...@peterbenjamin.com> wrote:
On Thu, 2012-04-12 at 11:51 -0700, Ariel Levi Simons wrote:

Your goal, for a sub-$300 air quality sensor kit, is very much along the lines of what we're trying to do. 

I have 12 Firefox profiles, and do not recall which one I used to find the Arduino compatible sensor boards (priced $30 and up).

I did do a quick look and found the raw sensors, very inexpensive, $5 and up.

   http://www.futurlec.com/Sensors.shtml

There is one sensor for multiple gases

   http://www.futurlec.com/Air_Quality_Control_Gas_Sensor.shtml

that needs associated pressure (or temperature) and humidity sensors to be more accuracy.  See point later below.

And looking quickly at one profile history, I found this with 70 air sensors:

   http://www.mouser.com/Sensors/Air-Quality-Sensors/_/N-55c0b/

most with circuit boards, running $30 and up.

More sensors

  http://iteadstudio.com/store/index.php?main_page=index&cPath=4&zenid=v9410sjq32j4u8d4pb46r1vpc1

None of these sites are the one I found with over a hundred sensors.


What the new group is about is to expand Safecast from just radiation mapping into a general environmental mapping platform. 

Terrific.  I'd like to see it be compatible with RadioShack's weather station Internet interface, to name one.  Both hardware and software and API.


My interest in this, besides capturing large scale environmental data, is to get students involved in carrying out research, especially in the citizen science model. 

Fantastic.


I'm a physics teacher who is building a student research program out at the Wildwood school in west Los Angeles

My BS is in Physical Chemistry from Harvey Mudd College, specializing in EMF and charged particles, with minor in EE, Management and Ceramics.  I'm now focused on immuno-neurology, nerve transmission, cosmology, and a few others I track.  Teaching and lecturing I try to do monthly.  I'm told I'm good with children.  Adults I find are slow.  ;-)


( http://scienceland.wikispaces.com/PEER ), and I've already had a few students work with Safecast.
Could you meet at CRASH Space this coming Tuesday at about 6:)0?  I'll be there either way from six to ten.

I'll give it a try.  I'm trying to close new contracts, so my time can be unpredictable rushes to prep the deal's paper work.

I like meetings with agenda items, and everyone leaves with some 'homework' they committed to research, implement, write, etc, within a time frame.  My days in college and aerospace showing through.  :]

Agenda Items

List of sensors needed on base array.
Application set of sensor types listing 2 or more sensors as addon to the base array
Mix and Match methods:
- Common high voltage supply 5 to 10KVDC for those sensors that need HV.
  -- Point is multiple high voltage supplies on the base circuit board is expensive
  -- List such sensors along with brand and needed HV.
- Time slicing 2 or more sensor types?
  -- reducing the overall wattage is desirable for solar or battery run unit
  -- Power up only one sensor at a time, wait for readings to stabilize, and move to next one
  -- Needed "alarm" levels moving towards dangerous levels
   --- Able to power up all sensors option
Sensors needing 'clearing' with fresh air issues
Sensors that get clogged issues
Sensor calibration, troubleshooting, replacement?  Kits?
New Items - suggestions during the meeting

Add your agenda items

These are technical implementation issues.  There are also human factor ones, that we should list.

Items for discussion this summer:

Los Angeles Grid areas (prevailing downwind of industrial strips, freeways, ocean, airports, tar pits, power plants, oil refineries, ... more...)
Central Processing or decentralized?  Crashspace pay for the electricity and bandwidth of a computer?
Backup of data locations?
Analysis power?  3D data visualization?  Supercomputer facility (It's free)?
Kit or prebuilt?  Both?
Design cost remuneration or not?  Rate?
LLC?  NFP?  Both?  Use someone else's?

I'll write up minutes and distribute them.  And prepare the next meeting's agenda list.

Peter




--
Ariel Levi Simons

Jerry Isdale

unread,
Apr 14, 2012, 7:50:29 PM4/14/12
to safeca...@googlegroups.com, crash...@googlegroups.com

Can this be streamed or on a g+ chat?

Although I'm technically in LA right now its just a flight connection back to my island. I miss crashspace.

- Jerry Isdale
pls excuse typos. sent from phone w dumb autocratic. err autocorrect.

--
You received this message because you are subscribed to the Google Groups "Safecast Air Quality Discussion" group.
To post to this group, send email to safeca...@googlegroups.com.
To unsubscribe from this group, send email to safecast-air...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/safecast-air?hl=en.

Ariel Levi Simons

unread,
Apr 14, 2012, 7:54:04 PM4/14/12
to crash...@googlegroups.com, safeca...@googlegroups.com
Jerry,

It might be kind of kludgy, but I can stream it through my g+chat account.
I'll setup a hangout then at 6pm Tuesday.
--
Ariel Levi Simons

Jerry Isdale

unread,
Apr 14, 2012, 8:13:58 PM4/14/12
to safeca...@googlegroups.com, crash...@googlegroups.com

Crashspace has a ustream channel. That could b good for viewing. G+ for making occasional comment or question.

- Jerry Isdale
pls excuse typos. sent from phone w dumb autocratic. err autocorrect.

Reply all
Reply to author
Forward
0 new messages