Bridging technology with growing(automating as much as possible)

[chowdan]

Hey everyone!

I am a very small time grower. I grow for personal use and do not sell my produce. I mainly grow for the please and the learning side of things in terms of gardening.

My 9-5 day job is with the IT industry, specifically in software testing side of things. I know some programming but am by no means a code monkey however I did just get hired as a SDET(software development engineer in test) for a large coupon company. I will likely learn Java, Javascript, Postgresql, SQL and a few other languages due to the new job requirements.

Just like learning any language, you can't read a french text book and learn french. You also can't learn french and never use again and expect to know it.

Programming is the same way, you must continuously use the langauges to keep up with it and be able to write programs.

Being a gardener not just for cannabis, but having an aquaponics based window farm in my living room, i have decided to figure out how can i go about automating this?

I have seen systems in use that automatically distribute out nutrients and keep water in a certain parameter, add water, and a bunch of other things. These commerical systems do not come free and are quite expensive(think blue labs dosetronic)

There is a lot of data we can collect from an automated system such as what nutrients are needing dosing and at which rates, what the soil/water temps are during day/night, specific light input the plant is receiving per every hour, how often one is needing to water, just to name a few.

Not only can you personally collect this data, but using a device such as a raspberry pi to be your "brains" of the unit would allow oyu to have a wifi connection to a server database.

there seems to be quite a few projects out there for this style of automating, however nothing really has taken off yet.

There is GardenBot which is ardrino based, which lacks the ability for data collect/wifi submission.

Has anyone persued this idea for cannabis or any other plants? I'm thinking if you had a database, you could completely optimize the growing for one specific plant over time as you would have a bunch of data points that you can create an algorithm to use.

Seems like a fun little project to pick up and explore. Hopefully by end of this year ill have a functioning prototype and be ready to build prototypes for others to "test" for me.

 

[stoned-trout]

automation leads to laziness which leads to problems....lol I wouldn't mind automation especially if it will raise lights too...

 

[chowdan]

automation leads to laziness which leads to problems....lol I wouldn't mind automation especially if it will raise lights too...

very good point. Though the other point would be its one less area that you need to focus on as much which gives you time to focus on other aspects of growing?

Either way though, i do agree it will lead to different issues.

 

[StoneyK]

The problem isnt working automation, its when it fails.
It can and is in many cases more work not only do you have to keep vigilance and maintain the plants but the machines that maintain the plants also need maintenance...lol
Equipment like this has a tendency to go down at the worst possible times too.
What if its a double wammy and the server fails and doesnt report the failure? Then your problem compounds.

Nothing and I mean nothing beats a master grower who works his crop everyday and knows whats goin on because he has eyes on it everyday.
I say keep it simple, a lil hardwork is good for the soul!

 

[chowdan]

The problem isnt working automation, its when it fails.
It can and is in many cases more work not only do you have to keep vigilance and maintain the plants but the machines that maintain the plants also need maintenance...lol
Equipment like this has a tendency to go down at the worst possible times too.
What if its a double wammy and the server fails and doesnt report the failure? Then your problem compounds.

Nothing and I mean nothing beats a master grower who works his crop everyday and knows whats goin on because he has eyes on it everyday.
I say keep it simple, a lil hardwork is good for the soul!

Hey,

Thanks for that, definitely great insight and helps me think things through on how this would work. Obviously its not going to be "build it, and it will work from the start" sort of system. Testing is everything in software/hardware. The more you spend in testing, the better product you'll create.

Yes technology has the tendency to fail, however so does the human body.

Depending on how i were to build this system, it really wouldn't be much work for maintenance. When you build a computer, it requires maintenance via software side of things, but not hardware. If you were to have a moisture meter within the soil, specifically, 4-8 meters throughout the pot, you can not only detect on when it would be best to water, but be able to gather data on watering.

The data you can grab from the moisture meters alone can be:
- How much water is being lost/taken up every X minute/hour/time
- How long does it take from the start of watering to end of watering to meet the required soil moisture level
- How often are you water?
- All of the above for each stage of the plant and specific species of plants.

This data alone will let you gauge more than "on/off" cycles for water, but will allow you to piece a database of knowledge for possibly detecting problems that are starting to occur, well before a human can detect.

The whole system can be monitored via cameras, every time something occurs, you trigger a image to be taken and mailed to X email address/posted to X location.

Yes there is the chance of something like a server going down without giving a proper error to the root user, however the likely hood of this is so minimal.

The world revolves around technology these days, its a matter of how you build out the infrastructure to prevent possible notification failures from failing.

Failsafes are a must, however having a camera(webcam, raspberry pi cam) is a simple method for you to be able to see your garden anywhere and at any point in time.

 

[MJPassion]

The world revolves around technology thesr days...

I'm of the opine that... Technology revolves around the world.
&
I have to agree w trout about technology making folks lazy & w Stoney about the added maintenance of the automated systems.
However, I also agree that once automation can be tuned and proven to be reliable, it would free up time for other mundane tasks like trimming... Also able to be automated but beat up buds are for consumers who don't care how their product came to market.

You might begin by making a list of the teks already abailable to this end then listing additional items you feel can be automated in your particular grow setup.

Also... keep in mind that the software really only serves one purpose, well, maybe two... Data collection & instructions of how to react to specific data points that are OOC (Out Of Control, for those that don't know).

My background is kind of in automation within a manufacturing environment. All I can say about it is that when the automated parts do not work, the factory shuts down completely until automation issues are resolved.

Even if you cannot get you project to work to the extent you would like, the data collected could still be used.

 

[beta]

.

 

[Goldy]

Hey man, I was thinking about doing this myself not long ago. Was going to go arduino based just because there are so many fairly cheap and useful add-ons. Pretty much everything the best controllers have and more. Prob the same as Pi just haven't looked into it.
Haha looking at what I wrote down - the possibilities are endless and its more an issue of narrowing down what is really, truly necessary and what are just interesting additions. But being able to control things to a tee without even having to go into your room could be very useful - think going on holiday or working away from home for extended periods. The ability to check on things via secured webcam and temp,ec,ph,CO2 readings etc.

 

[Goldy]

furthermore, for people on a budget, as the sensors are all individual modules you can add bits as you go. So start with just a basic temperature and ph monitor. And then add EC, timers, relays for lighting, power-backup etc. etc. Depending on your level of programming you could jump on ebay and grab a touch screen and make a neat interface.

 

[IntelliGeneS]

I know growth chamber controllers have been built using arduino.
Should also be possible to use both arduino and raspberry pi, I came across a soil moisture sensor setup which did that to allow web-based control and reporting.
Might mean you need to add Python to your list, but as I understand that's not a very difficult language.

 

[homebrew420]

Build it. Make it happen. The less of a slave you are to your grow the more you will enjoy the fruits of your labor. Automate everything. If one is checking pumps timers on the daily problems will be minimized.
#handwateringforthelose. haha
2k at the house and all is automated. Make the mistake of peeking in every few days to realize how much pruning is required. Good problems to have

Peace

 

[Daub Marley]

I used to think that upon legalization that all cannabis would be grown like more traditional crops in open fields, but after careful consideration I believe that automated greenhouses will grow most of the cannabis. Look at how the Dutch grow flowers and you'll get a good idea of the advantages and potential.