automating your grow lifestyle; on a linux box

[hysterix]

Hello everyone,

tl;dr - is there such a thing as open source sensors? If not is it feasible to reverse engineer sensors, or is it better to reverse engineer an already existing system? (growtronix for example).

q. hysterix, that is an epic wall of text down there, should I bother to read it all?

a. yes, only if you are bored!

I've seen there are a few automated grow room threads already, but I wanted to start this one as I am a bit more specific with what I want and will be doing here.

Ever since I figured out how to grow relatively well, I knew you could automate the entire grow. The only thing a human being would have to is physical moving of things, cloning, basic maintenance, etc.

I know there are some good things out there, and I have had my eye on growtronix.

I do not really mind the cost because of how capable things would become upon implementing all their sensors and things.

The only problem with that is now I am tied into the company growtronix and if they go out of business or something else I am out of luck.

I liked growtronix because they are using java as their platform so it can be run on any os, but of course truly mission critical apps should only be ran on a linux box as we want stability, reliability, and on totally open source software that is 100% free with no dreadful licenses to deal with. Currently no other os in the world offers all of the things I just listed, so it is easy to choose. Windows is unstable and has memory leaks, apple is a pay-for-support version of software you can obtain totally for free, that is instead tied to proprietary hardware. (mac itself is a hacked up version of openbsd linux) The concept of running this on anything other than linux is crazy.

The choice of flavor of linux is trivial, but for arguments sake, let's say this box is centos.

I want to run a full out os and not just micro-controllers because of all the fun I can have having it be on a real computer as I will explain later in this post.

The problem now is sensors. It is very difficult to organize the multitude of complex sensor data directly into this open source environment; there is always proprietary issues (from what I've seen) when you purchase a sensor that interfaces with a pc because they want you to run their software and that is not what I'm trying to do.

I'm trying to get the sensor, in all its glory into the pc, and then I can have my fun with it. From what I've seen, it appears growtronix will be as close as I'm going to get to this goal, short of ripping apart sensors, and figuring out how to dump data directly into the pc; how that reverse engineering is done on the hardware is a mystery to me, so I'm speaking with some assembly hackers as you read this.

I'm a software guy, and there currenly is such a thing as, "open source software".
Can any hardware vets. tell me why there is not such a thing as, "open source hardware". Why can't I buy hardware and dump the sendor data directly using a custom daemon or something else that is relatively easily. Why is it all sensor manufacturers want you to use their closed-source software?

Pretty much, if sensor manufacturer's were to give an api with the sensor that tells you how to interface to it, all would be well; I don't think such an animal exists.

I know there are microcontroller you can program with custom c, but that is very tedious and I know I could not do that myself, and I would not know how to go about getting the sensor data itself; again the hardware is a mystery to me.

Once I finally figure out this, I am planning on doing lots of fun things with this data. First off you can control this pc from within your own house, or anywhere in the world.

I stress that systems can be made very secure if you want them to be, you just need the money to fund the equipment and extra servers and it can be made pretty darn safe. So before anyone warns of the danger of having a grow room out on the internet that potentially someone could "hack" and control, know that could only happen if it is set up improperly, and/or a weak password is used.

The growroom itself is totally controllable anywhere in the world, from your own house or when you are on vacation. You can do anything from check different angles of the room, adjust your light height, check ppm and ph, balance ppm and ph, check other sensors just like how the growtronix system works.

I guess the question needs to be asked, why not use growtronix then?

Very simple, it is not 100% open source (to my knowledge) so I do not have full control over what is going on.

What I plan on doing can only be accomplished using open source, because I am going to incorporate a large number of other open source technologies and if done this way, it will be seamless.

Apart from the normal things the growtronix system offers in regards to sensors, I want to take it to the next level in regards to control over what you do with that data. I want to be able to create triggers for the sensors that do anything, call any script, send the data to do anything from anywhere.

What this means is that say your ppm starts falling out of range, and of course, the system normally is auto adjusting the ppm by injecting nutes depending upon a custom program created by you earlier and you adjust how the rez is dosed.

But what happens if something malfunctions and now your reservoir is not being dosed any more or being dosed too much? That is where the triggers come in and you can set them to do anything you want.

Want an email when the ppm goes out of range, you get one.
Want a phone call from an automated voice system telling you someone just opened up your window, done.
Want a text message your temps are abnormally too high, or the lights did not turn off or on when expected?

All this becomes possible when the code is open-source, and in proprietary systems it is not.

I hope I have conveyed this post properly. It is a hybrid question to any hardware engineers out there, and also just a statement of what I'd like to do. If you could not tell I'm a developer and I am going to combine my growing knowledge with application development.

I am currently in contact with assembly developers, and we are discussing whether or not it is more feasible to hack growtronix up to suite our needs, reverse engineer a solution using their sensors and write custom firmware for it; or try to do it all on our own from scratch.

I have had this subject on the back of my mind for many years, and I feel this is the appropriate environment to dump all this information and thoughts and ideas.

If anyone has any thoughts, ideas, opinions, advice, or just say, hey hysterix, you are nuts, I'd love to hear it!

 

[diyJunkie]

Interesting read.

I've often wondered why this hasn't been done before, seems like a good plan. I'll be tagging along to see how this goes.

However... I'm a mac guy - I want to see an iPhone app! I can see the commercial now - 'Need to check the PPM from the other side of the world? There's an app for that' ;)

 

[hysterix]

see diyJunkie I'm cool with anyone doing whatever anyone themselves want to do; but for me personally, I stay far, far away from the mac universe because apple always wants you to use their stuff (I dont blame them, they are trying to make money!). Want ipod? Oh wait you first need to download itunes! Oh wait you want itunes, okay first download quicktime! Want mac os? You need a specific piece of hardware; cant just buy anything you want from newegg!


In regards to cell phones in general I dont like them either. In fact I despise the whole telcom. system, it is ancient compared to what we could be doing already, and again, the only reason we are being held back is because it is closed source. (that and isps and telcoms are greedy)

What I'm getting at is I won't even own a "cell phone" (stripped down computer). Give me a mini laptop with a gsm card and I'm good to go; that way I can talk over voip on wireless networks without using any minutes.

Personally I think the best way in my opinion to get info to a phone is sms. I used to work as a developer for a company that set-up sms premium content. (stupid horoscopes to your phone) and I used to deal with "short codes" or those numbers you can text a msg to and get a unique msg back.

Pretty much we can set up the "short code" (they cost a hell of a lot) 420420 or whatever and you text PPM to 420420 and it'll text you back your ppm. :D

Don't be tied into any one piece of hardware, can be used on all phones and even e-mail; everyone is a winner. Just a thought.

 

[Titoon_29]

very nice read historix...

i ve been thinking about it too, but don t have enough knowledge in programming, nor time, for this...
i was just checking on growtronix, seems a very nice software...

i would try to build my own software, instead of trying to reverse engineering theirs... it doesn t seem to be a very hard software to program (not much sensors), and build yours will help you preventing compatibility problems... but you must know more about that than i do....

i remember using a software at school, which could easily deal with sensors, alarms, without any programming to know, it was a sort of visual programming... quicker, easier.... maybe you could use that for the user to easily add sensors or anything in the software...

you can go for linux, it even can be installed on mac now.....

you should first start to write the project specifications so you can have a good overview of the problems....

good luck !

 

[hysterix]

Titoon_29: the reason I say reverse engineer growtronix, is I am about 99% sure growtronix does not give out the api on how to interface with their sensors.

Conceptually I am going to "man in the middle" the sensor data and analyse how growtronix submits requests to each sensor itself and how it receives data back. Pretty much I'd be figuring out my own api.

I know growtronix does not give an open-source api to their sensors, because if they did they would not be very smart! That would be like giving away your business, so I understand why they do not provide this.

Because I see other people would like open-source sensors, I now know this is needed and it will be a future project. Setting up a web-site that gives different models of sensors, and the pre-made api to deal with each one. All this stuff should be free just like the os itself, and yes, I'm going to destroy everyone's business model that deals with grow automation (sorry growtronix) if I'm successful but I'm over it as I think everyone deserves to play with sensors in a free environment just like I do my code.

Hashtronaut: You hit the nail on the head, I did not go into it more because my post was long enough but that is exactly the point, logging. I'm all about data-mining, and the key here is storing ALL data the sensors and taking in so at any point you can go back and see why the grow 4 turns ago was better than the current one.

Once you get your 'perfect' (whatever that even means) set-up dialled into the system, you will pretty much just sit back and watch the dank bud grow.

Some people asked me if that would take 'the fun' out of growing, and I say to them not at all. It allows me to actually concentrate on the plants as opposed to worrying about trivial things like nutes, lights, and temps so, of course this system should only be for the experienced grower.

I do think this should ever be a way n00bs could set-up a system and not know how to grow but come up with dank bud; as a person first needs to understand what is going on manually, before they should move onto automation, in my opinion at least.


But yeah, the two main elements I want the most is being able to make adjustments to the room remotely, and perfect logging of the system so it is very easy to tell why one grow was better than the other.

I love graphs too, and with all this data it would kick ass to be able to see a line graph of your temps throughout the grow, laid over a line graph of ppm levels or something, so you can try to pick out similarities that normally you would have never been able to notice. (temps get higher, and ph spikes or something really crazy nobody has ever noticed before for example)

 

[Dalonewolf]

Taking care of plants cannot be fully automated. It is like taking care of children, thinking you can put them in front of the T.V and they will grow up to their full potential is a pipe dream (pun intended). that said you can automate alot and it can be helpful.

This is a starting point. The control anything link can be your beginning link the boards have A/D capabilities if your sensor requires conditioning then you will have to look at the other links and find a conditioner. all the information you need to get started is here and i feel that you have to look and research, I have my ideas but I'am not fully confident to give advice as to what steps for you to take .Having your mind set on linux before you have the facts can be a set-back and cause unneeded work. Remember most the work has been done your just trying to piece it together. E-bay has PH controllers that are surplus from manufacturing industries and go for cheaper than commercial ones for the hydro industry. GOOD LUCK









http://www.automatedaquariums.com/

http://learn.automationdirect.com/Sensors/Sensors.html


http://www.pacontrol.com/plctraining.html

http://www.what-is-net.info/what-are-solenoid-valves.html

http://www.eidusa.com/Interface_Boards_pH_ORP_ISE_420.htm

http://www.pacontrol.com/sitemap.html



http://www.eidusa.com/Interface_Boards.htm

 

[hysterix]

Thank you for the links dalonewolf!
Those links will keep me busy for a long time!

 

[darookie2000]

http://www.co2meter.com/collections...cts/cam-co2-meter-powered-by-auto-home-or-usb I've been looking at this one.

http://www.8051projects.net/downloads209.html this is based on a microcontroller, but it has all of the sensors other than co2

http://www.phidgets.com/products.php?category=6 a usb sensor kit

This one looks interesting, but the sensors look a little spendy

http://www.sparkfun.com/commerce/product_info.php?products_id=8311

http://www.pc-control.co.uk/greenhouse.htm I was reading this one today (it uses a package called digi-bee for USB IO)

PIC-based controller

 

[hysterix]

Thank you darookie2000; I knew there was a reason I posted this thread!

Keep them links coming!

:)

 

[baba G]

that's a long read...I didn't read all of it, so don't pounce on me if this question is deemed stupid.
Question: So, your in Fiji and your growroom wherever it is, is functioning properly. Next day your online and looking at your growroom to check it and all the sudden an exhaust fan stops working or something, how do you do anything about it???? Seems there is no substitution for having a buddy get paid to monitor things....truly automated is so, so so much easier outdoors after you set everything up. So can you do anything on a computer to help grow equipment fix itself...not yet! Just my .02

 

[darookie2000]

Thank you darookie2000; I knew there was a reason I posted this thread!

Keep them links coming!

:)

Yeah, I'm glad you did. I have definitely been looking for an automation solution for my project, and I've been keeping my eye out for sensor systems that I could grab just to see what they produce, then figuring out the automation part later. For me monitoring is priority 1, followed by automating things. I'd definitely be a lot less distracted if I could occasionally check a web page with all the measurements. Temp is my biggest concern right now, so I'd rather not run my desktop computer directly next to my tent, so I'm thinking of buying a netbook-type computer like the asus eee or the HP whatevernetbookmodelitis, though I have also found a computer that runs on like 10W or something.

 

[darookie2000]

that's a long read...I didn't read all of it, so don't pounce on me if this question is deemed stupid.
Question: So, your in Fiji and your growroom wherever it is, is functioning properly. Next day your online and looking at your growroom to check it and all the sudden an exhaust fan stops working or something, how do you do anything about it???? Seems there is no substitution for having a buddy get paid to monitor things....truly automated is so, so so much easier outdoors after you set everything up. So can you do anything on a computer to help grow equipment fix itself...not yet! Just my .02

This way, you could have the buddy go fix it, but he (or she :)) wouldn't have to be there all the time, just on an emergency basis

 

[baba G]

so, they would have to be on call, i suppose? I hear you , it is a workable idea for sure.

 

[hysterix]

Yeah, pretty much my idea of "total automation" isn't of course possible without an insane amount of money, and even that needs human beings to maintain the machines.

As good as you could possibly do pretty much without overspending too much money and keeping it feasible.

For example if an exhaust fan goes out, no amount of computer technology will fix that! A human being has got to get in there and fix it, this is for sure. At least though you would have an instantaneous notification that the temps are going up, or whatever bad thing happened, by phone or however you choose.

I guess you can consider this idea a very complicated warning system combined with a way to keep things running smooth by making adjustments for you automatically, combined with a lot of sensors that are able to be monitored remotely.

Absolutely people have got to get in there and clone, fix leaks, change bad pumps and fans, and of course obviously pests. Do not forget the main nute reservoirs would periodically have to be recharged also!

This idea though would allow a person to concentrate much more on the plants and not so much on the 'little things'.

 

[darookie2000]

so, they would have to be on call, i suppose? I hear you , it is a workable idea for sure.

Exactly