Chicago Eggs
Ed Borden
-
adam
Do we have a map of where the eggs went and what streams are associated in cosm? I know there is a Google map floating around, any chance someone can share that with me?
-
Javaun Moradi
Adam: I got on a plane while deployment was happening. I'm not sure if we have an up-to-date map of where sensors actually went. Can I send you a DM on twitter or a pvt email? I don't want to post the map the deployment team created on day 1 because I believe it contains home addresses with names.
-
Javaun Moradi
Also: how did we do yesterday? How many eggs deployed? Do we have a running list of Cosm feeds? Here is the feed for the official EPA sensor at 327 S. Franklin: https://cosm.com/feeds/68943
-
Robb Drinkwater
Adam: Charlie Catlett created a map, which I know Patrick had a link to. Javaun: I saw a list of several names associated with eggs. I have number 8, but am not sure that means that all eight went out the door. Alec: [Welcome!] As I understand it there were all pre-programmed to connect to individual feeds at Cosm. i.e. you do not need to associate your eggs to a personal account. However I searched on Cosm and could not figure out how to see feed data (Ed? Joe? help here?)
-
Robb Drinkwater
Update on enclosures: I checked mine this morning and discovered that mine did have some leakage. And as my Nanode was not on standoffs it appears there might have been some shorting around the voltage regulator. I'll be curious to see my feed (see above) and see when/if data collection stopped.
-
Alec Miller
I was tooling around on Cosm late and found it in the search https://cosm.com/feeds?q=aqe This looks like the list of all the eggs (also found the source on Github)
-
Charlie Catlett
If someone can send me a list of the final resting places for eggs I will create a safe map.
-
Robb Drinkwater
Charlie: I know of 5 eggs that went out into the wild; me, Bo, Collin, and Alec (two #3 & 5). Of those only Alec and I (#8) are posting data at the moment. I also know that four of the boards went with Raj for some of his trouble-shooting magic. When those return to SAIC working we'll dispatch them in new enclosures. As for making a map might I suggest we limit it to nearest intersection (e.g. I'm Augusta Blvd. & Rockwell St.)?
-
Patrick
I have a list of who took an egg. I'll post them up shortly, probably in about an hour. My egg, number 6, has been sending data since last night. However, it is indoors for the moment.
-
Patrick
My feed can be seen at https://cosm.com/feeds/69068
-
Patrick
I had a couple thoughts about preventing leaking... one is to use clear silicone (available at hardware stores; if you want it to be clear then make sure it's the 'dries clear' variety, which often is opaque white when it goes on) and the other idea is to just tape up a hat out of a plastic bag that would fit over the top component.
-
Robb Drinkwater
Hey Partrick: yup I see your feed is up. Also I should mention all egg holders (mother hens?) that if any of you need a custom length ethernet cable (i.e. personally I came up about 6' shy) just send me a measurement and we'll make one for you.
- Molly Finster
Charlie - Of the 2 we have, #5 is currently set up at our house and #3 will be set up at Argonne (as close to the weather station as possible). However #3 is missing mounting bolts and nuts - so that will be resady to go as soon as we can get to the hardware store. Might you have any ideas for a location here at Argonne that meets our requirements?
- Patrick
I will post a link to a public google map (likely the one that was already started) and all storks can mark their egg(s) on it.
-
Rajesh Sankaran
Yes, I have a bunch (4 ?) of them sitting in my car. I will test them in the next couple of days and try to debug them. I should be able to report back by the weekend.
-
Rajesh Sankaran
Alec: Wlecome, if you have not found your feeds yet, here are they - 5: https://cosm.com/feeds/69066
and 3:https://cosm.com/feeds/69064 . I am assuming 3 is not hooked up yet as it still has the
test value from yesterday, but 5 is doing great! - Rajesh Sankaran
Robb and Patrick: Do you mean the package is not water proof and you had some rain water leaking through?
-
Patrick
Rajesh: I didn't put my unit outdoors; I was just commenting on someone else's reported leaking.
-
Colin Hutton
I take it we didn't put the sealant on them properly then? I didn't take an egg because they were not ready when I ended up leaving.
-
Patrick
We didn't put sealant on them; that didn't just make it into the time frame. I think the sealing/water protection is going to be part of the collective beta-testing/troubleshooting phase we're starting on now.
- Patrick
The map that was shared with me yesterday has quite a bit of old information (prospective storks, etc) including personal information. To start the conversation about a group map, how about the following: I start a dedicated gmail address (I already did, actually: ChicagoAir...@gmail.com), create a public map, and invite all storks to put their eggs on it, cautioning them to be aware that all information on the map will be public. Any changes, addendums, or alternate proposals?
- Rajesh Sankaran
I think we have to seal it water tight such that rain water does not get in there, but not make it air tight / or create a air-pocket. I am wondering about the fan - forced air flow and I hope it is not forcing the air on the sensors, but making sure a stream of fresh air is available to them to compute the levels. I am sure the air flow properties of the casing has some effect on the measurements and has to be compensated.
- Patrick
Also, are there any mapping enthusiasts or specialists that want to curate or manage the map? The default will be that it will be connected to the dedicated Gmail address for the map.
-
Patrick
I'll wait a couple hours for comments, then post the map link up if I don't hear any changes to the plan or alternate ideas.
-
Robb Drinkwater
It was me reporting that I had a leak (after putting on my apartment window sill. See pic on Facebook) but thankfully the electronics took it. Wouldn't have been an issue if my board was on stand-offs (which it is now). I'm kind of against glues and sealants. If we can come up with a design that can be laser cut and doesn't leak (or allows for water) then we can ship enclosures to people 'flat packed'. [I'm working on some ideas]
- Patrick
Robb: cool on your alternate design efforts. I'm reluctant to use a gooey sealant because it won't look as pretty! :-)
-
Robb Drinkwater
FYI Joe: if you look at my feed https://cosm.com/feeds/69071
it looked like my CO sensor just crapped. Notice that there was a spike in humidity at about the same time; maybe the d.p. freaked? Gonna try a reset. - Robb Drinkwater
and FYI that worked. So we might want to examine the shield code and see if maybe we are going into some sort of 'lock up' state.
-
Javaun Moradi
Instead of glues, what about some kind of plastic laminate sheet (a single sheet) across the top of the case? Kind of like an iphone protector. You could even have a blue/green tinted one. Purty
-
Robb Drinkwater
Javaun: that could work. There are any number of solutions to this. It's a classic design problem and who better to solve it than us? Again, I am looking at a solution where either people can get to a laser cutter and do simple cuts from common materials or, if they can't, others who can will do so and send them out by mail.
- Patrick
Also, someone mentioned yesterday a compound applied to automotive electronics to protect them from moisture... anyone know what that is? My feeling is that there will always be some moisture in these enclosures, especially from moisture evaporating up into the unit off the surfaces they are sitting on, after rains and in the morning.
- Robb Drinkwater
That would have been me (why is it I'm always the "some guy said"?).
It's called conformal coatinghttp://en.wikipedia.org/wiki/Conformal_coating and is a way, usually using a spray on material, to protect electronics from moisture. But again not for your average user. I think most of the moisture problems can be mitigated through good design and intelligent placement of the electronics. - Patrick
Here is the map which is attached to AirQual...@gmail.com
I'll invite everyone who has an egg to the map; they in turn can invite others. Add your egg; use the nearest intersection if you do not want your address public, because...
THIS MAP IS PUBLIC AND WILL SHOW UP IN SEARCH ENGINE RESULTS. DO NOT PUT PERSONAL INFORMATION ON THE MAP OR MARKERS YOU DO NOT WANT TO BE PUBLIC.
- Patrick
putting the title of your marker in the format of 'Egg #6https://cosm.com/feeds/69068';
will allow one to easily copy the feed's cosm url without much clicking - Rajesh Sankaran
Patrick: Nice! I only wish one could get the Cosm feeds to float when one hovers / clicks on a position pin :). Well, if that is a long shot, it would be useful if the feed link is at least a clickable in the popup.
- adam
Thanks for the map! Excited to see it become populated.
I would love to start visualizing this stuff - unfortunately I am using "fake" feeds for numbers right now. -
Alec Miller
This is something from 'Captian Obvious' Don't put this thing in direct sun light. Since I set ours up at 10pm at night I wasn't paying attention and this morning when I was checking on it I noticed it the sun had poked out of the clouds and was beating down on it. I moved it about a foot closer to the house to get it into the shade and I think it screwed up the readings for the rest of today. I hit the reset button when I got home just to see if it will straighten it out.
- Robb Drinkwater
Alex, not as obvious as you might think; yesterday it occurred to me at well that the lid probably should have been mad out of white plexi, not clear, because basically we've made a little green house. *sigh* the problem is designers *love* clear boxes! We just just cut a few more today and a few more to do next week. We'll try white plex and if it seems the way to go we'll be glad to cut more and swap out for the current lids.
- Robb Drinkwater
Hey All, check this out: http://www.amazon.com/AIR802-POEPASS-01-Injector-Splitter...
(as suggested by Lindsay French, 2nd year ATS grad). Might be a solution for anyone having problems getting power & ethernet to an outside location. - Patrick
I still haven't had time to get mine outside... looking forward to it.
-
Patrick
Still only two eggs on the map at http://goo.gl/maps/e8Yy
... ping me via MeetUp if you need an invite to edit the map (I sent out invites to everyone that was on the board but can send additional ones). Also, I set the map up under a dedicated gmail account; we can all have access to that account to admin the map. Just ping me for the access info. - David Holstius
Egg #2 will be in downtown Berkeley when I head back to the office on Monday. Taking it around for show & tell, though, so readings will be sporadic
- Robb Drinkwater
Does anyone have a Nanode and sensor shield w/o an enclosure? I know Raj has 4 shields, but by my count there are two eggs unaccounted for.
-
Javaun Moradi
We really need lat/long for the Chicago eggs to get the most value out of visualizing the data. We really want to attach geodata to the Cosm feeds themselves, which are owned by Sensemakers (I assume Joe or Ed).
Please plot your egg here and we'll update Cosm with the geodata: http://goo.gl/maps/e8Yy
If you don't want to give away your home address, put it within a block or so.
-
Rajesh Sankaran
I have not gotten around to testing all of them, but I should be able to soon. I will let you guys know as and when I do. Thanks! Rajesh
-
adam
Yes, please please contribute to the map. Really wanting to visualize this stuff.
-
Patrick
If anyone needs to know how to make your marker on the map, just message me via MeetUp and I'll line you out on it...
-
Richard Beckwith
Robb said:
> FYI Joe: if you look at my feed https://cosm.com/feeds/69071 it
> looked like my CO sensor just crapped.Looking at the data, it seems like it's supplying some kind of "steady state" after having shown some variability before the freak out referred to above.
Anyone know what's happening?
-
Richard Beckwith
One thing to think about for new enclosure designs is to look at stevenson screens (google images, say). These were something that we had to build for an environmental sensor project a while back. We were told that the data wouldn't be useful for climate analysis unless we did. We just put a thermistor in one so that scale was way different so our design wouldn't be useful here.
- Patrick
Richard: Mine did the same yesterday (and continues like that today)...
https://cosm.com/feeds/69068 - Robb Drinkwater
Patrick et al. FYI I am no longer #8 ["I'm not a number, I'm a human being/"]. #8 is being passed off to one of the very deserving high-school student assistants who worked so hard making the shields. Charlie says he's away for a week but when he returns we'll make sure he gets on the map.
- Richard Beckwith
The numbers seem not to make sense. At cosm.com/users/sensemakers, inspecting data at one week interval, I see that AQE 1, 2, 3, and 8 have shown some variability but not much. The number associated with CO ranges from 188 to 627. Only 5 and 6 show much frequent variation. Has anyone tested things (like blowing smoke on it) to test for responsiveness? How many are out of doors? If you have one at home, could you please post some details? Still, this is playing out to be very cool.
- Richard Beckwith
Just went over the data with Adam. It seems like most of the eggs have reported data a couple of times and then the lines between are interpolated (used to be dashes on Cosm but now just solid lines). Only 5 and 6 seem to have been functioning consistently. It seems like 1 and 2 are reporting now, BTW.
- Alec Miller
3 was just assembled last Thursday night and was sent to Argonne on Friday. I tested it to make sure the standoffs didn't short anything out. I think someone powered it up on Friday for a short time. I have been having sporadic runs with #5. I have to reset it once or twice a day to keep the stat counters from sticking in one spot. It seems like it started having issues after the sun cooked it the other morning.
- Alec Miller
and one other side note about the flat lines with the counters for #5. My internet connection dropped one afternoon and it caused the readings to flatline until I reset the egg (I think it was Saturday afternoon). I don't know if there is something in the network code that may need tweaking.
- Joe Saavedra
So a few things:
First of all, let me say that this workshop was the most productive and inspiring on that sensemakers has ever held. So incredible to see so many people really genuinely considering problems and actually executing prototyping on several *extremely important* threads. This kind of energy is exactly why this project will be successful, and I need to thank all of you for that.
Robb and team did an amazing job of making sure we had the space and equipment to work so effectively in just two days -- i brought home one enclosure, and it seems incredible that it was designed and fabricated more than 10 times in so little time. Thanks so much, guys.
That being said, a few things:
Connection Droppings
- Network connected microcontrollers including both the Arduino Ethernet and the Nanode have this notorious problem of timing out after being online for one or more hours. All of the prototypes thus far have implemented this auto reset hack I wrote about here (http://weblog.jos.ph/development/arduino-reset-hack/), but we didn't get to implement this on the boards we made last week. Instead, we have a better method using the watchdog timer on the ATMega (Arduino) chip, but this was (my huge mistake) left out of the code uploaded to the Nanode.
I will be posting to GitHub here the updated code with the reset feature. I’ll post to the meetup when it’s pushed to the repo. http://github.com/jmsaavedra/AQE-Workshops/tree/master/July2012_Chicago
You’ll need an FTDI adapter or the FTDI cable to upload to code the Nanode (remember to select Duemilanove w/328) and then the dropped connection issue will be resolved. I’m so sorry, this is absolutely my fault that it wasn’t included on the flashes we did last week.
Sensor Readings
A few things about the sensors and what they’re outputting:
- All CMOS (metal oxide) gas sensors require a warm up time of about 1 - 12 hours before they begin to stabilize. This accounts for numbers that begin to rise or fall for sometime when they’re initially plugged in, and any funky values (sometimes they jump) in that initial day. (This is also a huge challenge when you consider going for mobile sensor units, as the sensors need to be constantly powered in order to give consistent readings -- something really hard to do when you’re working with a battery)
- The sensors are being read directly through the ATTiny chip that’s on the shield itself. What this means is that the ADC is reading the analog output of the sensors and then shooting it to the Nanode. What this comes down to is the following: All the numbers that are being posted are *entirely* dependent on the code that is on the ATTiny.
I will be the first to tell you: we pushed this version out a bit too quick. Vic (engineer at Lockheed Martin and Wicked Device ← AQE’s manufacturing and production lead) completed this version of the design just 2 weeks ago, and this was the first time we really put more than one together using both sensors and the I2C bus. The ATTiny design is absolutely the best way to go and will yield us the most consistent, reliable, and accurate readings. I’m extremely excited about it. And thanks (SO MUCH) to Rajesh, we were able to debug the hardware (yay cutting PCB lines) and get the individual sensor readings to the Nanode.
Rajesh and I tested the sensors outside with a cigarette, and the values began jumping immediately, and climbed for about 8 minutes. Walking back inside, they took about 5 minutes to return to the original values. So they are working, thank the gods. But this is a good time to point out the “response time” that metal oxide sensors have -- it can take a few minutes before the sensors output an accurately reflected value.
However, they’re not working as well as your prototypes *can*. Vic is actively working on updating the ATTiny code, and every one of your shields has the 6-pin ICSP headers so that you’re able to update that code as well. I’ll add this to the github as well when it’s iterated to a really good point (later this week). This one is a bit tougher to program as you need the Atmel programmer (Robb had one with him) to update. But i’ll post the hex file so you just need to install AVRDude and you can compile directly to the shield from terminal (no IDE necessary).
Again I want to thank all of your for your enthusiasm and huge effort. It was really inspiring, and extremely motivating. We’ll definitely be sending you PCBs of the next iteration which will be the final one.
Below is the list of the eggs out there. If you put a link to google maps next to yours, I’ll update the Cosm feed to reflect that location. Then you can pull the location directly from the API and map making will be super easy.
THANKS AGAIN, ALL!
- joe
1. Bo
2. David
3. Molly
5. Molly (for Argonne)
6. Patrick
7. Kathrin
8. Rob
10. Raj - Patrick
Joe: Thanks for an awesome update. I think I can speak for the group here in saying a general 'Thank You!' for a fantastic event. As far as any apologies over too-soon-code and so forth, you needn't worry one bit. The reality is that things never get this far at all when people worry about getting everything right first. In fact, one of the most insipiring parts of the project was the 'let's get these things built' leadership. Now that they're built, we can gradually improve them.
- Patrick
The process of getting it out and then improving it is the whole agile software approach in a nutshell.
As far as the map, what exactly do you want for mapping data next to our name in the google doc, as you mentioned in that same doc?
- Patrick
Richard: mine (#6) is still indoors
- Rajesh Sankaran
Status: AQE 9 and 10 are good, but the other two have some issue, most likely FW. I had to sort out a couple of soldering glitches and now will have to check if the Ard or ATMega is at fault. I will catch the I2C lines and that will give me some idea. Once I find the issue + fix the other two, I will send them over to SAIC. Joe, if the external ATMega's FW is at fault (not programmed), then I may need a AVR programmer and I will take you up on that offer :) if I can't source one here.
-
Rajesh Sankaran
FWIW, the two devices 9 and 10 were placed next to each other and you can the variance between the two. Most readings are within 10% of the other. While that is acceptable in the case of CO and NO2, I am not sure why the Temps are so varied. Joe, shouldn't the digital thermometer be accurate to less than 1% ?
Robb Drinkwater
Raj
Cesar Garcia
Good work guys!
It seems like you are on the express track of AQE deployment! I would suggest to check an email sent to the list about ten days ago to the list regarding Madrid workshops. We discovered there some clues there that appeared again during your deployment, like sun and aqes don't work well together, some enclosure design tips using recycled recipients and some trials testing 8 egg in the same place (under dcdcity username).
Best and keep the flow,
César
Adam Laskowitz
Rajesh Sankaran
Nice to see more sensors being added to the map. How are the others doing?
aim: rajeshxsankaran | voice/text: 225-328-1183 | web: http://www.mcs.anl.gov/~rajesh/
Raj
Adam Laskowitz
Rajesh Sankaran
It seems only 1 egg is streaming now?
Adam Laskowitz
Ben
Rajesh Sankaran
Joseph Saavedra
Victor Aprea
Joe,
I apologize, I haven't had the bandwidth to be involved with the Ethernet code much. I actually don't think a hack like that being proposed is necessary. The currently recommended library for the Nanode is EtherCard (github @jcw), and it handles DHCP lease renewal properly (which I suspect is the tour cause of these types of problems). If you are using EtherShield, that is no longer maintained. Migrating from one library to the other is not top too difficult either.
If I'm way off base, sorry... I'm writing this while on the road do I'm speculating without reading the code...
Regards,
Vic
Rajesh Sankaran
patrickaroo
Thoughts?
Jeffrey Munie
Is anyone up for a get together? I can offer our beautiful offices (Open Software Integrators) in the West Loop, along with some food and definitely beer. We could hack on some sensors, some code, and otherwise catch up on the project.
Thoughts?
--
--
Jeff Munie
Industrial Design
www.jeffmunie.com
www.vicariousdesign.tumblr.com
Alec Miller
--
Robb Drinkwater
~Robb