Community Developed Mods


#1

Hi everyone!

I spoke with Caleb last week in SF and one thing he mentioned that would be useful from the community are mods, which a number of us are already doing. Now that we have a number of FC’s up and running, I think it may be helpful for us to discuss the efforts on mods currently so that myself and others can have some direction on what would we could spend time developing/collaborating on.

So far I’ve come across the following mods:

  • Inexpensive cooling system
  • Constant immersion/inexpensive pH and EC sensors
  • Nutrient delivery automation
  • Nutrient analysis tool

Before diving deep into any one of these, what are your guys’ initial thoughts on these and other possible mods? What other mods have you considered? Are some worth pursuing more than others? Do any of you want to collaborate with each other on a mod?

I especially would love to hear what the team at OpenAg thinks so we don’t end up working on something that comes out in v2, for example.

Cheers!
Adrian

“Coming together is a beginning. Keeping together is progress. Working together is success.” -Henry Ford


#2

Indeed there is value in innovation :slight_smile:
Great to see the community in action. Without this forum at least my Food computer build would not have been possible.

I have considered and even ordered parts for the cooling mod. In case this is the nutrition cooling :slight_smile: Just because I found that cooling element awesome and wanted to see it work and the parts are affordable :slight_smile:

Sensor thing is a problem needing solving. At the beginning I thought that the sensors were going to be immersed into the solution and that would be fine. To be honest they are the most expensive bits of hardware in the build. So I am eager to see where this “mod” or bug-fix will go.

Nutrition delivery and analysis also interesting. Lets see where it goes.

On a more general side I would like to understand more what is it trying to achieve and what is the target of each in detail. I understand it is the awesomeness of building things that drives some of these things and I think that is plenty reason to go forward and build it.

But…

I’ll use the cooling mod as an example. And just off the bat I want to mention that I think it is a great mod and definitely one of the coolest :slight_smile: (notice the bun there) mod to do :slight_smile:
Aim is to make the nutrient solution cooler then the grow space air. There is a thread here
Inexpensive parts :slight_smile: working solution. Nice piece of mod log.
What I would like to know is that what is the data behind this? Is there a botany article I could read? What is the exact temp we would need to achieve with the cooling system? As a newbie to growing things I have a hunger for all data :slight_smile:
Also am I to expect that cooling the water temperature to X is going to give me increased harvest or earlier harvest?

I would like to suggest one more mod to the list :slight_smile: Simply add current sensors on each segment of the FC to calculate the cost of electricity. This would then log the energy usage of each segment. For example Grow lights, heater, humidifier and so on. We could then calculate the cost increase or savings of each mod.
Like if I replace the window with a solid insulation block… How much less heating is needed?

I’m also a big fan of modularity. In case I want a food computer with just the basics and then have the possibility to later add a nutrient cooler or a delivery analysis system without a complete overhaul of my FC.


#3

I agree with @silversson here that the community has been a great help to move forward. Really appreciated.
For now I have not been able to do a lot of testing but I think there are some improvements possible on the air box:

  1. not sure if air heating is actually necessary as I noticed that a some heat is still generated by 90 Watts of LED light.
  2. why not combine the humidifier AND the CO2 input into the air box and thus save 1 fan. (use the air box for circulating air in the grow chamber and add CO2, humidity and heat when required.
    CO2 sensor:
    as the CO2 sensor is one of the most expensive components I would suggest to, either come with a cheaper, yet workable alternative (see here the comments on the cheaper version ) or make an agreement with a single supplier to get discounts for the open source community.
    I agree (again) with Jussi that the window seems to be a huge ‘heat leak’. Maybe “double glass” could be a simple improvement.
    About cooling: is it a weird idea to combine cooling with pressurized CO2 dispension ? If you release liquid CO2 is will freeze it surroundings instantly and increase CO2 levels at the same time… Not sure if that is doable though.
    Looking forward to read other suggestions as well.

#4

I haven’t looked too deeply at the build, but what is there in the way of hardware fault tolerance? Given that the unit has to run for extended periods and the intention is that it’s reasonably hands off, are there any considerations given to pieces not working? Someone who actually runs one would be in a much better position to speak to it, but, for example, without a filter or regular cleaning, dust can build up on a fan (especially if this is in a garage) and cause it to burn out - would the computer know?

Redundancy is always expensive, but alerting the operator to some of the more common hardware failures (are there any?) seems like a good value add. Perhaps it’s more of a software mod, but might also require extra hardware to be effective.


#5

I’m assuming by this you mean analysis of the nutrient solution, not the needs of the plants? My understanding is that actually analyzing the nutrient solution is a giant pain - the sensors are expensive and need to be regularly calibrated, and they’re real “lab” equipment, so they need to generally be cared for. Given the relatively small consumption of the nutrients, has anyone done a back-of-the-napkin calculation as to where nutrient analysis makes sense, and where it makes sense to just flush the solution and start again with known concentrations? For the larger size computers, maybe it becomes economical, but I’m suspicious that it doesn’t make sense for the personal food computers - both in terms of money and time in maintenance/calibration.

I haven’t really looked at what solutions (pun not intended) people have come up with to mix, but a naive approach could be a means to flush the solution, a dosing pump, and a solenoid manifold. If nutrient reservoirs are stocked with known solutions, getting the desired mix is just solving a system of linear equations, which seems relatively straightforward.

Maybe a pain, but can you link to them? It’d be nice to compile some reference material for mods.


#6

My FC is producing its second batch now,
Where some part of the hardware could be improved ( indeed the sensor) I believe its UI interface is could be improved.

For instance

  1. when the activator are on or off a small led like should appear red or green in the icon
  2. Alarm, we should be abel to get alarm (by email ?) when the T° or the pH is below set-point in order to act on it
  3. historical trend, would be nice to to follow sensor ( T°,…) over time and over set point of the recipe, a little bit like a music card in the music automate.
  4. A display of the time clock of the recipe to know when it will be “day” , or “night”

Those are only some point but it would be nice…


#7

Good ideas! I would suggest the integration of a push service (pushover, pushbullet) and/or IFTTT integration (don’t know if thats already done).
IFTTT integration would be awesome, as it opens up a huge amount of features.
For example:

  • switch a philips hue light to a specific color to indicate ph/temp warnings
  • send mail on alerts
  • send a push notification with the current state of the PFC when I get home
  • create tweets on events
  • send images/post images online or save to dropbox (nice for a timelapse video)
  • turn WeMo power switches on/off

And much more :slight_smile:


#8

I’m not yet familiar with how the event/alert system works for the Food Computer, but I could definitely set up an IFTTT integration if someone can point me to the right direction (I work there).

These few questions might be a good starting point

  • Is there an API that provide either a log or stream of events/datapoints? Is it publicly available, or is it only available in the LAN the RPi is conneted to?
  • Are alerts, or data from the sensors being aggregated somewhere? or is everything is kept in the RPi?
  • Is there some sort of authentication system in place? (e.g. a website where you login to view your plant’s data)
    • If so, is it able to provide OAuth2 tokens?

This indeed would be awesome.

:slight_smile:


#9

Nice! Love the activity here. So here’s an updated list of all of the projects that are going around:

  • Module - Cooling
    **References: 1
  • Module - Air Exchanger
  • Sensor - pH
  • Sensor - Electrical conductivity
  • Sensor - Current
  • Sensor - CO2
  • Shell - Window
  • Automation - Nutrients Delivery
  • Tool - Plant Tissue Analysis
  • Tool - Electricity Usage Tracking
  • Failure Monitoring
  • “Basic” Food Computer
  • User Interface
    **References: 1
  • Push Service integration (e.g., IFTTT)

I attempted to add a bit more order to the project list. Let me know if you guys have any suggestions of how to describe these. If anyone is interested in talking about any of these in depth (e.g., problem definitions, design criteria, design alternatives, etc.), feel free to create a separate forum topic to better capture that discussion.

If anyone has any more topics to reference for the mods, please post them for compilation.

Also, does anyone have experience collaborating on open source hardware projects? I know this book exists and could possibly serve as a good reference. I think it would be helpful for us to learn from others’ successes and failures.

I think we’ve got our work cut out for us :slight_smile:

Thanks for all of the responses everyone!


#10

I talked to Caleb about this mod. I think he was talking about a plant tissue analysis tool… @Caleb do you recall which mod this would be?


#11

I’m not sure how feasible it is, but in my opinion, it would make sense, to replace/modify the top camera module to allow the image based analysis of the plants i.e. hyper-spectral imaging (HSI). Found a research paper, where a research team hacked a DSLR camera.
Optics is not really my area of expertise but for the sake of not forgetting about it, I figured I’d share it here :slight_smile:


#12

Plant tissue analysis tool? That sounds interesting. Is there a way to do this that’s not invasive? The only nutrient analysis I’ve seen has required harvesting a bunch of the plant matter. Tough to do when you’re only growing a couple in a personal plant computer, maybe more worthwhile at the larger scale.

I’ve seen leaf turgidity sensors, probably not very in a hydroponic environment but likely useful in an aeroponic grow. I agree with your sentiment @adrianlu - it’s cool to gather all everyone’s ideas in one spot. Lots of different perspectives.


#13

Great book!


#14

We are also experimenting with using cameras or Kinect as a way to measure plant “leafyness” and estimate biomass (thread).


#15

The mod that would help me the most is air cooling.

Here in Brazil, even inside houses, it can get quite hot. Then what happens is that the vent fan is always on, since its effect on temperature is set as negative. But because of that, the humidity and CO2 controls become very ineffective, and neither of the three variables reach their set points.

I built a cooling module based on Peltier modules similar to what MrGadget did in his nutrient solution cooler, with a cpu water cooler on the hot side and an aluminum heat sink on the cold side, with a fan blowing air on it (or away from it). I tried with one, two, and three Peltier modules in parallel and couldn’t get the temperature to decrease a single degree. I think one module doesn’t have enough power, but two or three should work, even a little. But since it didn’t, I think the problem might be that the “dissipation” of cold to the air is very poor with forced ventilation. I saw that most people only use these modules to cool water or attach them directly to an object.

I’m starting to think on the possibility of using a used AC or to take off the cooling system of a mini fridge like this guy did.
http://www.homebrewtalk.com/showthread.php?t=317662
But I don’t think these are the best solutions for the PFC.

If anyone has some ideas, I would love to hear them.


#16

LED development.
Hi all, yesterday I was fortunate enough to be attending the AVF (association for Vertical Farming) summit 2016. It was in Amsterdam, for me just around the corner…
Amongst other very interesting discussions I remember a debate on the LED technology to be used for verical farming. The general conclusion was that ‘broad spectrum leds’ (white light) is probably the better choice compared to blue-red technology. The reason for that was:

  1. health care for employees (probably not an issue fur us food nerds…)
  2. there is a lot of yet unknown processes inside a plant that require light (beyond photo-synthesis) that might influence the taste and ‘feel’ (crispiness) of leaves. This process could be influenced by other wavelengths than red or blue.
  3. having ‘full-spectrum’ means that you actually have, at least 3 colors of leds. This gives you the option of influencing the intensity of various wavelengths during the various life cycles of a plant (germination, growing, blossom etc) which might improve the quality of the output.

Alternative Hardware for DIY Food Computers
#17

Hey all,

I updated the list. I think we’re gathering some useful information here. Now, what do do with it…

Module - Water Cooling - does not exist in current version
**References: 1
Module - Air Cooling - does not exist in current version
**References: 1
Module - Air Exchanger - Goal?
Sensor - pH - constant immersion and lower cost
Sensor - Electrical conductivity - constant immersion and lower cost
Sensor - Current - measure electricity usage
Sensor - CO2 - lower cost
Sensor - Camera/Image Processing - measure plant “leafyness” and estimate biomass
**References: 1, 2
Shell - Window - reduce air and thermal exchange
Automation - Nutrients Delivery
Tool - Plant Tissue Analysis - tool to allow lower cost and faster analysis than current methods
Tool - Electricity Usage Tracking - how much energy does v1.0 use?
Hardware Failure Monitoring -
“Basic” Food Computer
User Interface Improvements
**References: 1
Push Service integration (e.g., IFTTT)


MVP - Future of Process
#18

I want to bring this old post back up - here are some potential projects for those of you with the V2 PFC up and running:

https://wiki.openag.media.mit.edu/contributors/rob.baynes/projects_for_anyone

Inexpensive cooling system - Recommendations for Water Chiller?
Constant immersion/inexpensive pH and EC sensors - Still a big issue, we’re exploring using the camera to log PH based on a test strip instead to avoid PH sensors altogether.
Nutrient delivery automation - This is one of the projects outlined above
Nutrient analysis tool - Something like this I think would be a great way to start digging deeper into the water chemistry: http://www.urbanaquaponics.com/content.php?147-Aquaponic-System-Testing-Test-Kit-Reviews