Greenhouse Kickstarter Data Collection Stretch Goal

Paul is wondering what "Really awesome data collection" would look like for a stretch goal on the kickstarter.  Many of you know much more about this than me so I'm just getting the thread started and throwing some amateur ideas in the pot to get started.  There are many data collection systems out there that could be discussed thoroughly but I think here we just want to figure out what the dream system would do.  What data it would collect and how it would deliver it.  Let's try to avoid the rabbit hole of which exact sensors and dongles and data plans and all that stuff for now.

I think some awesome data to collect would be:

outside temp

inside temps (ceiling, north end above beds, foot level, cold sink level, bottom of casing(s)

soil temps in 4-5 places moving from one point above the membrane to one point 6" away from the logs of the greenhouse to show the temp gradient

soil temp deep in the mass (if that isn't a part of the bullet above)

temp of grey water entering the system?

temp in the middle of a log in the ceiling, wall and cold sink

temp in the greywater planting bed

humidity of greenhouse air

solar light received

ounces/gallons of greywater entering the system per hour/minute

I think it would be splendid to collect it on a 10 minute or shorter cycle and upload it to permies for anyone to look at.  Or at least store it and grab a month's worth of data at a crack and then upload it.

Please add your thoughts or suggestions!  Here's one quote I mined from another thread that touched on the subject:

Derek Callihan wrote:I haven't had a chance to listen to the podcast, so if this is already in the works, average minds sometimes have the same thoughts as great minds? Throwing an idea out for a stretch goal: Built-in instrumentation at multiple levels. Thermometers at each level (Roof, Beds, Trench, Bottom of Pipe (if you can manage it)), humidity gauges, light meters at multiple locations, etc., all tied into an automatic logging software that monitors things 24/7.

Combine this with things such as having multiple well casings and destratification pipes that you can open or close and you'd be able to get good, hard data on if and how much of an effect each has, ideally looking to find the point of diminishing returns for each, and how they interact with one another. Additionally, this experimentation could be extended to things such at the color of the back wall. Start with unfinished wood and get your baseline measurements. Then try painting it a bright, reflective white. Then maybe try a reflective metal foil cover. With the instrumentation in place, it makes quantifying any differences from these changes much easier to illustrate and allow you to more quickly spot any potential problems.

To understand how the system is behaving I drew up this system boundary pic in case it may be handy for thinking about what to measure.  We have significant energy flows via radiation in and out of the glazing, in the greywater, in water within the envelope (vaporization and condensation phase change energy transfers), via conduction in the earth masses logs and glazing, via convection in the air phases....are we missing anything?


We monitor each of these so as to understand the rate and magnitude of energy flow.  You have done a great job of capturing a lot of that in your list.  Here's are a few others I'd add to allow us to understand how all the bits are functioning:

-air in destratifying pipes, temp (to understand flow driving force) and flow rate (to understand how well air is being pulled into the casing)
-soil temp at 20', near casing and far from casing (measure the distance between the two....these measurements set the gradient that allows us to calc the rate of energy flow when adding or removing energy....maybe save some of that soil....can measure thermal conductivity)
-condensation rates....flow rate might be hard to measure unless collecting that water to send back to the Abbey then it's temp and flow rate could be measured as an energy out flow


Any other system components we want to understand the functioning of?

Nice Greg!  My only new thoughts are:

I think 80-90% of the heat loss will be due to conduction and radiant losses through the glazing.

Measuring flow rate in the pipes will be really hard.  Low pressure flows, I believe, are a real struggle to measure with any amount of accuracy.

I assume the amount of greywater coming into the greenhouse will be small, unless they install a shower soon in the Abbey  Come to think of it, where does the grey water go after it's processed?

Might we want to measure moisture levels around various logs in the wofati structure?  To see if they stay dry enough to be structurally stable for decades?

I think the biggest things we want to learn from this experiment(s) are:
1.  Does the casing system work?
2.  How well does the umbrella/membrane work?  What's its actual R value.
3.  How much effect does the cold sink have (independent of the well casing)?  This would further prove Mike Oehler's awesomeness.
4.  Do the logs stay dry now that they're also touching a humid greenhouse?

Hi Mike, I know you didn't want to jump to meter specifics yet, but in general terms I just wanted to say that mass flow meters using heat exchange between a sensor and the air flow stream can measure as low as something like a hundredth of a ml/min flow rate and I don't believe they create much of a pressure drop in doing it.  Past bed time now, will revisit when more awake.  Thank you for this thread.

Sweet, I didn't realize there was such a thing.  Thanks!  I shouldn't've started to say what could and couldn't be done.  This is brainstorming and the money is unlimited (at least for 28 more days)

Maybe two outside temperatures?
One in front, as in in Greg's diagram, to measure the effect of the structure in creating a micro-climate in front of the glazing.
Second behind, like the snowman in Greg's diagram, to get a true ambient temperature.

Plus it would be redundancy built-in in case the one poops out mid-winter.

The main barrier to us doing this kind of data collection is equipment availability and compatibility with our constraints. We have no wifi on the Lab, so anything that requires wifi is out; many/most weather stations that we are interested in have required a wifi gateway. We need the data collection to be somewhat automated (not something you have to go plus a USB drive into to get the data) and the data to be conveniently shareable, preferably with visualizations, in order for it to do much good. A bunch of text posts in a thread with me manually transcribing temperatures from a screen readout once a week is not going to be of much value. And we need the equipment to function to at least 20 below zero F for the outdoor stuff--which most tracking thermometers don't--and/or to be waterproof and tolerant of humidity depending on location. This kickstarter might open up new doors in terms of equipment, since it could expand our budget significantly compared to when I was looking for tracking thermometers for the Abbey last summer, but so far I haven't been having much luck. I am open to suggestions, however.

One of those satellite based weather systems that tracks precipitation, air pressure, wind speed/direction and temperature and satellite links it to a web based platform that displays the data. I think it would be useful for a lot of the experiments being done.

Bonus if it could display the rest of the data being collected real time as well with pretty graphs. Ooh ! A camera that automatically takes daily photos would be neat too.

Interior data with humidity, temperature in multiple points.

One of those gadgets that takes the infrared temperature of surfaces, and someone who goes and measures the temperatures of several set points within the greenhouse weekly.

Oh, and as a huge stretch goal if you wake up with a couple extra tens of thousands... installation of a thermistor to measure the ground temperature every 1-3' below the greenhouse and below another adjacent spot.

Jennifer Richardson wrote:The main barrier to us doing this kind of data collection is equipment availability and compatibility with our constraints. We have no wifi on the Lab, so anything that requires wifi is out; many/most weather stations that we are interested in have required a wifi gateway. We need the data collection to be somewhat automated (not something you have to go plus a USB drive into to get the data) and the data to be conveniently shareable, preferably with visualizations, in order for it to do much good. A bunch of text posts in a thread with me manually transcribing temperatures from a screen readout once a week is not going to be of much value. And we need the equipment to function to at least 20 below zero F for the outdoor stuff--which most tracking thermometers don't--and/or to be waterproof and tolerant of humidity depending on location. This kickstarter might open up new doors in terms of equipment, since it could expand our budget significantly compared to when I was looking for tracking thermometers for the Abbey last summer, but so far I haven't been having much luck. I am open to suggestions, however.

I haven't had a chance to watch the Youtube video from Friday yet, so apologies if some of this has already been covered. I think to get a good handle on the starting point for instrumentation, we'd need to know the constraints we're working with. You say there's no wifi at the location. Does that mean there is currently no hardwire connection and/or there is no practical way to get a connection there? Is there any cell reception at the location?

Is there any electrical connection there and/or is it practical way to run electricity there (purely for instrumentation). Or will everything need to be run on batteries? If batteries, could there be a possibility of a small solar array and battery station to split the difference and allow the instrumentation to be wired?

Derek,

There is no hardwire connection and no way to get one. Recently, AT&T put in a tower and we can finally get cell service up there; we are planning to buy an AT&T phone and use it as a hotspot to facilitate some of the livestream stuff we are offering as rewards for the kickstarter. But we are not planning to keep paying for that monthly plan once the rewards are delivered.

There is not really a practical way to run electricity there consistently. We have a solar array but the lead acid batteries are shot and the inverter pulls a lot of power, so it basically only works when the sun is shining, not consistently and not at night. There are about a hundred hours of conversation to be had about the solar cart and batteries, but bottom line is that we're switching to more plug-and-play lithium battery banks for reasons. And the ones we have will supply a few hours of power before needing to go back to base camp and get charged. So basically anything we use needs to be battery powered.

Hope that helps!

Could the equipment be powered by a lithium battery bank and a dedicated solar panel sitting next to the greenhouse?  Then the equipment could run off A/C power if needed and the battery wouldn't need to be recharged at base camp.  Along with avoiding the trouble of getting the sensors and equipment to not start blinking when the battery goes for a charge.

As an idea, at my allotment site we have no permanent electricity. So we have a 30w solar panel, deep cycle battery and the data collection all runs off a raspberry pi- they're 5V and relatively low power. We have about 30 ds18b20 temperature sensors, as well as humidity sensors (htu21d, or hih3000s I think), light sensors (bh1750), and a few others (there's a barometric pressure sensor in there somewhere). These aren't high end calibrated sensors- they're a few dollars each though, so it depends on how much accuracy you want.
We have the pi generate its own wifi, then wander up with a phone/laptop and download the data to peruse at our leisure. But you could also save data to an sd card/usb stick to swap out.

Well, at least until we hear officially, let's assume that we can include a solar panel and battery as part of the "Data Collection Stretch Goal".  

Charli has a good proposal for one way of collecting data.  From what I know of it, those sensors are quite affordable.  Setting up the Pi and knowing how to do the programming is a slight challenge but I'm guessing a boot or ant up there will know how to do it.

Are there other systems that could work with a small source of continuous electricity (battery/panel)?

What ways are there to store, retrieve and/or send the data somewhere?

I'm rather mechanically minded so if anyone wants to take over this conversation and ask better questions, feel free

I can help you with the setting up and programming- I'll even gain and program the pi here and send it to you if you'd like A pi zero is less than £10, and the sensors are in single digits. Or I can teach someone what to do- I've done quite a few systems like this now (including my own greenhouse).

Other possible systems- arduino. very similar but even lower power consumption. Can't be its own wireless, so you'd want a non-wireless way of getting the data (sd card/usb stick) or also run a router for the wireless off of the same power supply)

Or picaxe- even lower power consumption, can run these off some rechargable AA batteries and a tiny power supply but they're harder to program (in my opinion). Can save to an sd card using openlog easily enough.

Thats all the chips I've ever worked with!

Seeing as how it looks like the electrical issue has been tackled, I'll give a go at attacking the internet connection issue. It looks like AT&T isn't nearly as generous with their offerings as some other companies, but it does have some. Looking at them, rather than getting a phone, a wireless hot spot would give you more flexibility at a lower cost. The two viable options would be either the Global Modem USB800 for $180, which is a USB device, so it's limited to one device, or the Nighthawk LTE Mobile Hotspot Router for $250, which is a wireless router connected to a wireless hotspot, so you can connect multiple devices. Both of these would require a data plan from AT&T, which only offers two plans for them: $50 for 10 GB a month or $70 for 15GB a month. 10 GB a month is probably overkill for just uploading data, but that's your lowest option. Looking at the reviews, it seems like the biggest potential problem, assuming you have a strong enough signal, is the rather limited range of the wifi.

Just my opinion, but setting up a dedicated hot spot at the lab (assuming the greenhouse will be in close proximity) seems like it would be adding a lot more options to future endeavors. As long as there is power at both locations, adding a pair of network extenders ($30 each) should be pretty straightforward. A little back of the envelop math and it looks like you could cover about three acres with that set up. Adding up the costs, assuming the router, a pair of network extenders, and the lowest data plan for three years, plus a bit of fluff because there's always the unexpected costs, it looks like you could have internet there around $2500.

On the instrumentation side, it sounds like Charli is the man with the experience and knowledge. It looks like the Pi is the way to go, being able to collect and log all of the different inputs from the variety of sensors and wirelessly transmit them out. I'm assuming it would be fairly straightforward to setup a database that automatically extracts the data, organizes it, and generates some pretty graphs to look at. Having that database linked to a webpage accessible for backers (or anyone) to look at in more-or-less real time would be an added benefit that seems like it would take minimal effort to maintain once it's set up.

WiFi for local use MIGHT be solved by using WiFi dongles. I know that older Raspberry Pi boards use these to connect. I don't know if they would work in a computer. But, most routers, sold within the last 5 years, would have built-in WiFi.

Cabling with Ethernet wire could be done done depending on the length of cable needed to gain access.

OR, you could simply buy a cheap Raspberry Pi computer ($35 to $50) with a small Vid-Screen and place that in, or near, the greenhouse to collect data and store it on a large micro-chip that could then be carried to your office computer.

I've attached a PDF File that shows schematics and configurations of equipment to monitor greenhouse operations. You could add weather types of add-ons and other devices, all very low cost and reliable, to do monitoring of temps, water temps and water levels, etc. Check it out!

Something that I should have noted in the Post just above is that all of these things in the Garden Monitoring System can be run on a 12 Volt battery with Voltage Regulators. I believe the input Voltage on the Raspberry Pi should be at 5 Volts. There should be 12 V Pumps available at reasonable costs. Many of the other hardware pieced mentioned for monitoring run on top of the RPi boards as "Hats" (add-on items) and pull current from the RPi board.

SO, maybe a small Solar Panel would keep the one 12 Volt battery required charged up to keep everything running properly.

I love this idea, having played  around with quasi passive structures for a few years now. I do have a couple questions (hopefully I’m not being redundant with another persons inquiry):

My first reaction to the design drawings was, “that seems like it will be very cozy in terms of temperature, but will it provide enough light for optimal plant growth?”

Have I missed some research on how little light many subtropical (like those that need greenhouse in MT)  plants really need? I know a lot of shade loving plants would do fine with dappled partial sun, but it looks like mid summer sun would only reach the windows for a couple hours on the shoulders of the day. I think most subtropical fruiting plants would benefit from more light than this. So just some observational data on such plants’ (ie tomatoes) growth, and whether they get leggy or stay nice and stout would be greatly edifying. Thanks for putting in the work!

I guess I just ASSUMED that everyone, as I do, has electricity in their greenhouses for lighting.

Winter growing, even if heated, would require a good selection of cold hardy plants any way as those are normally very early Spring plantings and there is less light during those early weeks of growth.

I would assume that all Brassicaceae and other early Spring plants like peas should do well with less light. Also the radish family and lettuces should grow well too.

Other than using the electricity to run something like full spectrum or other specialized LED lights to extend the daylight hours I don't know how the growth would go with the short hours of sunlight and some heating ability.

Ben Zumeta wrote:but will it provide enough light for optimal plant growth?

Ben, Jesse that subject is being discussed on the thread about the greenhouse design.  For example, this post from Josiah, but you can read the surrounding discussion as well.  

There is the possibility to change the floorplan during the design phase, especially if the Kickstarter raises enough money to increase the size of the greenhouse (one of the stretch goals--also discussed in that same thread).

Ben's comment made me realize that it would be nice if the (Raspberry Pi?) data logger system had some ambient light sensors as well.   Then it would be convenient to graph the actual light received at a few key locations together with the temp, humidity and other variables.  Jen wouldn't have to remember to write "it was cloudy today..." somewhere, snow-reflected-light would be accounted for, etc.   Just an idea to stir into the pot.

Edit:  Of course, Mike already said that in the first post of this thread.  Doh! He also said:  

here we just want to figure out what the dream system would do.  What data it would collect and how it would deliver it.  Let's try to avoid the rabbit hole of which exact sensors and dongles and data plans and all that stuff for now.

My weather station (“Ambient Weather”) has a watts/m2 reading. Here it varies between 900-1100 watts in full summer sun, which is about ideal in my experience with indoor grow lighting.

A BH1750 will give you lux readings, its a cheap i2c sensor and easy enough to read from.

It's ok Kerry, about three days ago I suggested we start talking about sensors and how to store/distribute data.  So feel free to carry on!  

I just checked about data collection from sensors to the Raspberry Pi and it looks like data can be ported directly into a spreadsheet which can be opened using Excel or OpenOffice - but, it would take some coding to do that. IF, IF this is going to be a Project, maybe some of those young people in the Boot Camp crew know Python to be able to code the necessary inputs for the spread sheet to gather the data in a neat package that could then be turned into graphs.

Here is a site that details some of the processes required to do the coding/porting to get the data saved directly into the spreadsheets.
      https://www.instructables.com/id/Data-Collection-With-Raspberry-Pi/

ALSO, the PDF that I added to one of the posts actually shows some of the basic coding needed to run some of the devices in the Automated Gardening System.

Charli Wilson wrote:...at my allotment site...the data collection all runs off a raspberry pi- they're 5V and relatively low power. We have about 30 ds18b20 temperature sensors, as well as humidity sensors (htu21d, or hih3000s I think), light sensors (bh1750), and a few others (there's a barometric pressure sensor in there somewhere). These aren't high end calibrated sensors- they're a few dollars each though, so it depends on how much accuracy you want.  

Charli, that sounds great!  30+ sensors off of one RPi?  I was wondering whether some/all those sensors are far apart from each other, connecting back to the RPi with cables?   Can the various sensors be 10+ meters from the Pi (in different directions) like they would need to be in Paul's project?  Is that what you are doing there?

Does the Pi have any kind of display?   Does it ever need any kind of maintenance?

Charli Wilson wrote:We have the pi generate its own wifi, then wander up with a phone/laptop and download the data to peruse at our leisure...

What is the data-file like that you download from the Pi, and then how do you use it on your project?   Is it maybe a CSV file that you then open in Excel or something like that?   Or something completely different?


What you have done for your allotment sounds very close to what this greenhouse will need.  I'd like to hear more!

Currently my house pi has 34 sensors, and my greenhouse 29! Some up to 70m from the pi. They're mostly DS19B20 temperature sensors in a 'one wire' configuration (actually uses 3 wires), using easily available cat5 (computer networking) cable. You can either put all the sensors on one wire and circle it around- or with a bit more kit (which isn't expensive) have several wires going in separate directions.

I don't have a display, as in a screen on any of my pi. However the pi itself has a web server installed on it, and provides its own wifi. So you can wander up with a laptop/smart phone, connect to the pi and see the data displayed in a nice table- I tend to show current temp and min/max in the last 24 hours. I also have graphs of various things- there's some examples on this thread here: https://permies.com/t/40/58216#1083338
It would be possible to have a screen plugged into the pi to display the same thing- either an old computer monitor or a specialist little pi-screen.

The downloaded data you can have in any format you wish! I store everything in a database and tend to generate the graphs straight from it, but no reason you can't download it to a csv or an Excel file to play with later. I've got 6 years of sensor readings stored on less than 1GB of space in a MySQL database, but other flavours are available.

Leave it with me a day or two and I'll find the instructions I wrote when I made everything, and include/link to them!

[quote= I tend to show current temp and min/max in the last 24 hours. I also have graphs of various things- there's some examples on this thread here: https://permies.com/t/40/58216#1083338
It would be possible to have a screen plugged into the pi to display the same thing- either an old computer monitor or a specialist little pi-screen.

The downloaded data you can have in any format you wish! I store everything in a database and tend to generate the graphs straight from it, but no reason you can't download it to a csv or an Excel file to play with later. I've got 6 years of sensor readings stored on less than 1GB of space in a MySQL database, but other flavours are available.

Leave it with me a day or two and I'll find the instructions I wrote when I made everything, and include/link to them!



HI CHARLI:   It sounds like you have an ideal setup. I'll keep an eye out for your forthcoming information on the system.
I have a RPi3 with a camera module but haven't used it for some time now. For a "monitor" I couldn't find anything small when I went shopping for one in BestBuy, but, they did have a small CHEAP television, very lightweight that I use for a monitor. It works great.

I haven't tried to download data into a spreadsheet, but I do want to use the camera module and download pics into my computer, or maybe another RPi with a huge amount of storage - as I want more of the camera modules around running on the RPi Zero WiFi units for data collection. In essence, a security system.

[quote=
Winter growing, even if heated, would require a good selection of cold hardy plants any way as those are normally very early Spring plantings and there is less light during those early weeks of growth.

Other than using the electricity to run something like full spectrum or other specialized LED lights to extend the daylight hours I don't know how the growth would go with the short hours of sunlight and some heating ability.

I checked out the flat panel concepts for heating - both air and liquid - and I think they both have problems when in a very cold climate. For one, a long-time heating process just to heat up the panels, as well as the air and/or water. This would be a handicap if you are trying to pump in warm water into several 50 gallon drums, to last through a long night. I think the design I had in the PDF file would still be the ideal method - quick heating of the unit and flowing hotter water into the drums for the short time ideal sunlight is available. And, yes, this method also has problems, one being the possibility of overheating and melting down. But that could be controlled with an automated cover, and if automated (computer controlled) the computer could also run a module to direct and turn the mechanism to the best possible position for maximum sunlight.

Jesse Glessner wrote:
HI CHARLI:   It sounds like you have an ideal setup. I'll keep an eye out for your forthcoming information on the system.
I have a RPi3 with a camera module but haven't used it for some time now. For a "monitor" I couldn't find anything small when I went shopping for one in BestBuy, but, they did have a small CHEAP television, very lightweight that I use for a monitor. It works great.

I haven't tried to download data into a spreadsheet, but I do want to use the camera module and download pics into my computer, or maybe another RPi with a huge amount of storage - as I want more of the camera modules around running on the RPi Zero WiFi units for data collection. In essence, a security system.

Check out motioneyeos- an easy to set up pi-based CCTV system. You can use the onboard pi camera, ip cameras or seperate smaller-pi based cameras. We use this setup at the allotment as well!
https://github.com/ccrisan/motioneyeos/wiki

So we usually budget $500 for a stretch goal. If y'all want to put together a budget for $500 with links to products that meet the criteria I outlined earlier in the thread, we would really like to include this as a stretch goal. The $500 would also need to include a reasonable payment for whomever was taking on the data collection task based on how much effort it requires (more if you have to record it manually, less if it syncs automatically, etc.)

The baseline budget includes four tracking thermometers/humidity sensors like the ones in the Abbey:

https://www.amazon.com/SensorPush-Wireless-Thermometer-Hygrometer-Android/dp/B01AEQ9X9I/ref=sr_1_1_sspa?dchild=1&keywords=SensorPush&qid=1594758038&sr=8-1-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUEySDJYN0pITjlIUURKJmVuY3J5cHRlZElkPUEwOTA0MDM2N0k0Qk5JTE9GMzhPJmVuY3J5cHRlZEFkSWQ9QTA3MzMwNzgxUTBWT1BTV05IMTdOJndpZGdldE5hbWU9c3BfYXRmJmFjdGlvbj1jbGlja1JlZGlyZWN0JmRvTm90TG9nQ2xpY2s9dHJ1ZQ==

So no need to include those. I think the next highest priority would probably be probe-style thermometers for soil temp (in the mass, for instance). But that's not set in stone. And if there's going to be a battery bank or solar setup, that would need to be priced in.

I have been talking to Paul, and he is willing to spend a bit more if we make this a milestone stretch goal (maybe $150,000).

I'm thinking something like this:

A tracking thermometer in each of the well casings (4 total).
One at the bottom of the trench.
One midway up the trench.
One at ground level.
One at roof level on the back (mass) wall.
One at roof level in the middle of the greenhouse.
One probe-style thermometer buried in the mass.
(So about $500 worth of equipment at about $50/thermometer)
Once every two weeks, someone gets paid $100 to record and post the data, plus do some microscopy to compare the water coming in to the water going out and report the results. ($1,200 total, $100 per session, 2 sessions per month for 6 months, and then a $1,000 bonus for completing the test = $2,200).

We already have a microscope, but if we could figure out what kind of camera attachment thingamabob it needs and buy one of those (unknown cost), that would up the coolness factor of the data collection a lot, since we could post pics/video of the microscopy.

So $2,700 + whatever a microscope camera attachment costs -- say $3,000 total cost for the whole stretch goal.

Thoughts?

I'm hoping Charli will chime in on this.

If we mimic Charli's system we could have dozens of temperature sensors in all the places you suggest and they'd cost under $5 ea.  Every sensor for an arduino or raspberry pi (that I think we'd need) are generally under $20 ea including the cable.  So my hunch is that we could get the sensors and brains all done for under $300.  The tricky part is the data storage/sharing/uploading.  That's where I know nothing.  But I'm guessing the other experts on this thread can weigh in with lots of good info.

C'mon team, you can do it!  Help Jennifer out!!!

So I can more easily keep track of what will be a fair few posts, I've started a new thread to document me making a raspberry pi temperature data collector, and it is here: https://permies.com/t/144157/permaculture-projects/Raspberry-pi-temperature-data-collection

Jennifer Richardson wrote:I have been talking to Paul, and he is willing to spend a bit more if we make this a milestone stretch goal (maybe $150,000).

I'm thinking something like this:

A tracking thermometer in each of the well casings (4 total).
One at the bottom of the trench.
One midway up the trench.
One at ground level.
One at roof level on the back (mass) wall.
One at roof level in the middle of the greenhouse.
One probe-style thermometer buried in the mass.
(So about $500 worth of equipment at about $50/thermometer)
Once every two weeks, someone gets paid $100 to record and post the data, plus do some microscopy to compare the water coming in to the water going out and report the results. ($1,200 total, $100 per session, 2 sessions per month for 6 months, and then a $1,000 bonus for completing the test = $2,200).

We already have a microscope, but if we could figure out what kind of camera attachment thingamabob it needs and buy one of those (unknown cost), that would up the coolness factor of the data collection a lot, since we could post pics/video of the microscopy.

So $2,700 + whatever a microscope camera attachment costs -- say $3,000 total cost for the whole stretch goal.

Thoughts?

My thoughts: Is the instrumentation no longer wrapped into the $100,000 stretch goal? $150,000 would be beyond even the most optimistic projections based on the data from the previous kickstarters and it seems a shame to place something that would add so much more information and value for almost nothing (1.5%-3% of the total raised) behind a stretch goal that is unlikely to be met.

Has there been discussion of the power supply and internet issues and have those been built into the calculations of the stretch goals? For this thread, we were assuming that a power supply in the form of solar panels and a battery bank would be available to power the instrumentation. Has there been any discussion towards providing internet connectivity at the lab and/or greenhouse?

Looking at the setup Charli put together in his thread, it looks like you could have the instrumentation and recording devices assembled for around $300. I personally would add include humidity and light sensors to provide solid answers to the people questioning the effects those things will have in the winter. If you wanted to add a bit more of the "WOW" factor, include a camera, or three, to take photos at set interval, such a every 4 hours during the daylight hours. Putting an analogue thermometer and hygrometer in the shot will along you to cross-reference your instrumentation readings as well as being able to have a visual record of everything that occurs. As a bonus, it makes it super easy to make cool time lapse videos showing the progress.

Based on some very quick math, I believe that would run about $500. That leaves $2500 of the proposed $3000. From my post above, it appears you could have internet connectivity at the lab/greenhouse for 3 years for that price. This would make logging and uploading the data much simpler, as it could be automated, and eliminate the need for someone to manually handle the data logs, which makes up a large majority of the cost currently. Not to mention opening to door for the many additional opportunities that would be available with an internet connection there.

On the microscopic analysis of the water, I think that should be a separate stretch goal, as it seems independent to the instrumentation. I would also question the usefulness of the information you would obtain from doing it. If someone is more knowledgeable, please correct me if I'm wrong, but I believe all you would be able to see it that the water coming is has more particulates than the water going out, which can be observed visually. To have an accurate measurement of how much cleaner the water is at the end, you would need a turbidimeter, which is a bit of a sensitive and pricey instrument.