Grow Automation and Industrial / Computerized Systems Controls

[SRGB]

LargePrime: We need VAtransplant here.
What if we really did select one board and stitched all the needed parts together to do an OSS/OSHW grow room controller project?

Hi, LargePrime.

That might still be one of the primary goals of the (Float Valve Automatic Watering) project. We have been terribly distracted of late, relevant to contributing to the details of the project. The schematic at the last page of the linked thread should provide a general outline though, for further development.

pi, beagleboard, aosp, or from scratch should be doable. Interestingly, there was a recent web commercial from a cellular carrier about how they were `helping` farmers by developing/deploying systems to help farmers monitor water, etc.

The main item would be determinging the limits of the application(s). What exactly would the gardener want to control?

There might be a number of ways to accomplish the task. From simply using wireless switches to turn gear on or of, to fuzzy logic systems. Just designing the interface could be an entire project itself. A lot of the work has probably already been accomplished via space exploration programs from over a decade ago.

There might be a great deal of items that contributors should possibly consider if the concept is truly intended to be open source, or `free` hardware/software. The first being, perhaps, do you the contributor truly want to contribute and release open source, or free software? Clearly defining those terms, clearly defining any attached licenses. Also, bearing in mind that some that may view the open source documentations might then turn the end-product, or parts of the contributions into commercial products. What restrictions will be attached to the project? And, if those restrictions are exceeded by an entity, will there be an effort to maintain the open source nature of the project? That might be an entire task in and of itself - to monitor if items developed by the team are filtered out and used in commercial products. That might be ok for some, not for others. A contributor could get an offer for six or seven figures from some industry or the other. A contributor might sincerely want to profit off of the final product, etc. Perhaps that is one reason some teams for `non-profit` entities to manage such matters, as it can be overwhelming to simultaneously develop gear and be attuned to whether or not the internals of the gear are being used in commercial products for profit.

Perhaps slightly off-topic, though currently relevant, it might be illuminating to read about the beginnings of `twitter`, and the apparent tumultuous history of that team, up until the present. Briefly, some of the initial contributors might not be as close as they started out.

In any event, it might be fully possible to do the work here, aware of the above cursory details. ICMAG is an open forum.

The modules should be scalable and configurable. Though many differ on defining scale; for some, it might be to garden for the pure joy of it, for others farming might be for their livelihood. The modules must be reliable for both the micro gardener and the farmer depending on the seasons` bounty to feed their families. For some, the module might only need to deliver nutrients or water, while others might need a full industrial grade system, with remote access capabilities. The latter might be relatively straightforward to develop, with a single focus; for the latter, validation of data, networking, backup power systems, etc.might be required for failsafe operation. Again, these points go back to clearly defining the scope and goals of the project Then, the modules have to be continually tested. In short, it is a lot of work. Some might desire some form of return for their work, weighted by their contributions. All of this should be considered by the prospective contributors before the work commences.

There are probably some members here at ICMAG that are far more aware of plc`s than we might be. We would probably focus on easily acquirable and configuarable components. pi, odroid, or even repurposed handheld (already having board, sensors, arm architecture) for initial hardware with aosp might be some interesting platforms to work with.

Again, nice thread. +K.

Best,
/SRGB/

 

[StitchMonkey]

YES! This IS new, just like the advent of the internet. This is giving us a much finger level of control over all the devices that operate to make our lives more computable or give us things. This kind of having machines do our bidding will grow exponentially, and controls such as these are the necessary link to make them work for us. You raised the hair on the back of my neck...

It's good to hear that sensors are cheap, motors already have multiple speed points for relays that all these systems will operate, etc.

I just got the feeling that I found the Lego box I'll be playing in for the rest of my life, LOL

The funny thing is this attitude is party why I am so happy about it getting fully legalized in my state as well as Washington. This is not just fun for us to geek out about, but frankly someone out there is going to make money off of this as well! Not all of us want to build this ourselves, just like there is a market for pre-built/custom build gaming PCs, there will be a market for prebuilt systems for this, and frankly I love that state is well poised to benefit.

It is a huge opportunity. And just like anything else we will borrow from other industries and they will borrow from us. My guess is as people start doing this more, and talking about it openly and improving things, other areas of household control will start to also seem like good options. As we get better at managing the environment of grow rooms, our tips and tricks can be applied to the house as a whole, as we get better at automatically managing intake and exhaust and redirection, we keep indoor environments more comfortable for less money/energy (in theory at least).

For instance as more nerds get more comfortable working with ducting and venting, I would not be surprised at all if we start seeing water cooled PCs that exhaust the radiator into an exhaust vent that can then be routed either out a window, reducing AC need in the summer, directly back into the home heating system, redirecting the heat out of just one room to be more evenly distributed in the winter, or even vented into a small grow area for warmth. How cool would it be to grow some seedlings heated from your gaming?

 

[Ttystikk]

We need VAtransplant here.
What if we really did select one board and stitched all the needed parts together to do an OSS/OSHW grow room controller project?

Yes, we need him. How do we send an invite?

I think doing the project you suggest is brilliant! I benefit by building a system and using my facility for testing, others contribute code, ideas, whatever.

The funny thing is this attitude is party why I am so happy about it getting fully legalized in my state as well as Washington. This is not just fun for us to geek out about, but frankly someone out there is going to make money off of this as well! Not all of us want to build this ourselves, just like there is a market for pre-built/custom build gaming PCs, there will be a market for prebuilt systems for this, and frankly I love that state is well poised to benefit.

It is a huge opportunity. And just like anything else we will borrow from other industries and they will borrow from us. My guess is as people start doing this more, and talking about it openly and improving things, other areas of household control will start to also seem like good options. As we get better at managing the environment of grow rooms, our tips and tricks can be applied to the house as a whole, as we get better at automatically managing intake and exhaust and redirection, we keep indoor environments more comfortable for less money/energy (in theory at least).

For instance as more nerds get more comfortable working with ducting and venting, I would not be surprised at all if we start seeing water cooled PCs that exhaust the radiator into an exhaust vent that can then be routed either out a window, reducing AC need in the summer, directly back into the home heating system, redirecting the heat out of just one room to be more evenly distributed in the winter, or even vented into a small grow area for warmth. How cool would it be to grow some seedlings heated from your gaming?

I AM in Colorado, not that it matters if I'm developing technology to reuse waste heat, that kind of tech can be used nearly anywhere.

I AM extremely interested in this system's possibilities for integrating growroom control, home HVAC, reuse of waste heat for heating and other uses (greenhouses to hot tubs!), water chilling and heating, even outdoor aquaponics and using the tank the fish live in as part of the climate management system.

There is no doing this with a patchwork system of timers, thermostats and single room controllers, so I'm at the point where something like this is required to both handle what I've got going now, and what I plan to build out in the near future.

In addition, the idea of controlling my greenhouse/growroom, lawn sprinkler (enhanced with spent hydro nutes!), house, pool and hot tub temperatures and whether I need the snow melted off the driveway all from the comfort of my living room easy chair sounds pretty darn twenty first century chic to me!

 

[LargePrime]

Hi, LargePrime.

Hi, SRGB.

The main item would be determinging the limits of the application(s). What exactly would the gardener want to control?

I was thinking something similar. perhaps in phases? Yes we have Ttystikk who wants to automate the world, but perhaps we should start smaller? Then perhaps evolve that in phases? Start with a minimal, perhaps noob, or basic growing system/layout, and look at automating that? Then grow (ha) from there into systems that support more common/complicated grow scenarios/more icmag users? Perhaps recruit victims?

There might be a number of ways to accomplish the task. From simply using wireless switches to turn gear on or of, to fuzzy logic systems. Just designing the interface could be an entire project itself. A lot of the work has probably already been accomplished via space exploration programs from over a decade ago.

Lets not jump to tech until we figure out the actual problem. And i imagine large parts of what we will use will be recycled from other people. I am pretty sure there are already OSS grow systems already, which we should look at.

You might have an idea about what a system you would want should do. I might have a very different set of ideas.
Then each of us building our own systems is one thing.
Building a system that both of us can use and work well with is a whole other beast.
ANY system that just about anyone can come and use is a whole other thing.

Clearly defining those terms, clearly defining any attached licenses.

Free, as in beer! And we do not have a choice.
Anyone posting anything on a public free forum cannot be expected to control said thing, at all, in any way.

This community shares tips and techniques to do things better. If someone uses that info to make a bunch of money, what can we do? What should we do? I posit the answer is nothing and nothing.

The REAL strength of forums like these is that IDEAS can be subjected to a form of peer review. And better ideas will be sorted from lesser ones.
My hope is any grow tech we come up, if any,will follow that same process.

Perhaps slightly off-topic, though currently relevant, it might be illuminating to read about the beginnings of

That is what happens when you have an owner. I don't see this project having an owner. Do you?

In any event, it might be fully possible to do the work here, aware of the above cursory details. ICMAG is an open forum.

Caveat Contributor. lol

The modules should be

YES, yes, a thousand times yes! Except that parts where people get anything at all. The beer is free. They can take what they are given or help us make better free beer. Or get lost.

There are probably some members here

YES, yes, oh... wait. Used that already.

Again, nice thread. +K.

same to you sir! And best to you as well.

 

[LargePrime]

Yes, we need him. How do we send an invite?

It seems you could PM him a link, as you have over 50 posts.

 

[LargePrime]

wait now i do too, so i can PM him too!

 

[LargePrime]

PM sent

 

[rives]

Something that you may want to give more consideration to is the overall architecture of your system. Basically, there are two approaches, both of which have their good points and their bad. From what I gather from your previous posts, you are looking at a centralized control scheme, wherein all of the I/O (inputs and outputs) go to one central processor.

The other option, and my preference after doing decades of system automation, is distributed control. This uses more specialized components to do the interface with select portions of the system. Some information is relayed to the central processor to communicate system functionality, and the central processor can take steps to bypass that control if necessary, trigger alarms, or whatever is appropriate.

For example, I mentioned above that I intend to use a PID loop temperature controller for my fan. This has the option of using 0-10vdc or 4-20ma output, digital temperature displays of the current temperature and the setpoint, programmable alarm functions, and compatibility with a variety of sensors. This will be set up to vary the speed of the exhaust fan to maintain a given temperature in the sensing zone, and if for some reason things get out of range, an alarm will tell the PLC to override the analog control and go to a fixed speed. If this doesn't pull the temperature within the desired range, then the PLC can start adjusting the lighting to lower the heat load by either dimming the lighting or shedding some fixtures, or a combination of both.

This could obviously all be handled in the PLC (or at least in one with analog capabilities), but the problem that I've seen in the past is that if the PLC quits, if you want to alter the program, etc, then everything comes to a halt. I like redundancy, and I like having options. If you get the control of your system too tightly focused, you can very easily lose both.

 

[LargePrime]

Something that you may want to give more consideration to is the overall architecture of your system.

redundancy is very important for system robustness.

But right now we don't even have a system. While chained cheap controllers are a great idea, I would like to see us just get to the point where we have a working one first.

 

[rives]

Approach it however you like, but the system architecture becomes the foundation for all further component choices. If it isn't chosen up front, there can be a great deal of time and money expended in developing capabilities that won't be utilized in the end result.

 

[LargePrime]

If it isn't chosen up front,

see, i was not listening.


Proposed Spec #1, chainable boards.

Pros;
Long term system growth and robustness. Redundancy. expandability, flexibility

Cons
more work verifying and testing.

rives, how would you see this working?
If we go by your example, and your controls are assumed under the control of the child board, and heat is out of control, and all mitigation has failed, what should happen?
How does a chain-able set of controllers help?

also for all your review
http://socialcompare.com/en/comparison/low-cost-arm-boards

now i imagine any page like that to be out of date, but one board i was looking at was this one
https://www.olimex.com/Products/OLinuXino/A20/A20-OLinuXino-MICRO/open-source-hardware
notes are 2 core, 1 GB ram, and 160 gpio, so it can control EVERYTHING

off to research chained controllers.

 

[rives]

rives, how would you see this working?
If we go by your example, and your controls are assumed under the control of the child board, and heat is out of control, and all mitigation has failed, what should happen?
How does a chain-able set of controllers help?

It would depend on the circumstances as indicated by the various sensor inputs. If the sail switch in the air stream tells you that you still have air flow, then you can assume that the fan is functional but can't keep up with the heat loading. With my setup, I can either start dimming by increments and watching for a temperature drop over a reasonable period of time, or dumping light fixtures and waiting for a result.

If the sail switch (or pressure switch, or motor current-sensing relay, or whatever) tells you that the fan isn't functional (the motor gave up, the PID controller went gunnysack, etc), then the PLC can take control of the fan motor with a control loop in parallel to the PID controller and feed it with a fixed speed signal. If this doesn't bring up your inputs in short order, then the fan can be assumed to be dead and you either shut the lighting down completely if your system can't handle any heat load without ventilation, or you drop to a maximally dimmed setting and start rotating through the ballasts so that you preserve the lighting period and still evenly light the grow area.

Optimally, this would be accompanied by sending a text message to your phone to tell you that there were troubles, as well as plugging the information into an event logger so that you would know that it happened.

 

[SRGB]

LargePrime:

Lets not jump to tech until we figure out the actual problem. And i imagine large parts of what we will use will be recycled from other people. I am pretty sure there are already OSS grow systems already, which we should look at.

Hi, LargePrime.

See reprap.

Free, as in beer! And we do not have a choice.
Anyone posting anything on a public free forum cannot be expected to control said thing, at all, in any way.

Well, beer is not free. Even if you make your own; ingredients, the time to make sure the batch goes right, doesn`t go stale, flat or express an odd taste. Even if your buddy hands you one after work. It has to get produced by some means. But, we do get the rub. However, as one continues learning in life - during the process - one also might become aware of the finite nature of `it`; or, rather, the value of time. Some might take this very seriously, and won`t give you any of their time without receiving prior compensation. Not posting that is our approach, just that time and energy are probably the the most valuable `things` one might `have`. The above noted, we do `get` the rub and where you are coming from. Thanks for sharing!

Also, there are moderators on public boards, perhaps for good reasons. It might be that not all who lurk or post do so for purely benevolent reasons.

While one may do what they whilst, some might respect certain items more than others. Even the creative commons share and share alike type licenses ask that attribution be displayed in a conspicuous place, next to the content. We will not go into the entire range of `IP`, but it has perhaps occurred where a tech is revealed, later to be remixed by others, without reference to the original tech. Neither here nor there, but if a team works hard on a project, while it is true that they cannot `control` others` responses to the tech, they can `manage` the tech itself; whether for posterity, or simply accuracy.

As a cursory example, some time ago we did some research into open source licenses. Again, if one intends to do the same, perhaps have some `time` on your hands and don`t expect everybody to agree simply because something is labeled `open source` or `free`. Both the MIT license and the BSD license are both considered `open source`; we believe the former permits any person to remix, repackage and sell the released work. The latter, at least the `3-Clause` version forbids the use of the `names` of the creative work from being used as an endorsement of the creative work.

For example, a hypothetical `automation` team here endeavors to make sure everything is `free` on the boards, the final build is released. Some huge VC sees the opportunity, reads the tech, and simply outsources production of the tech, contact the `major distributors`, and voila, new product on the market based on the hard work put in by the team. `Free` for the VC with resources, not so much free for those that put in long hours figuring out bugs in the system, documenting the bugs, releasing fixes, version control, etc. We are not here trying to post one way or the other on the matter, just that that side of development - of any new `thing` does exist. To omit that from consideration in the preliminary stages would probably not be advantageous.

An actual example might be the several `defensive patents` that exist relevant to certain naturally occurring compounds in certain species of plants - as having medicinal value. Not necessarily by ordinary inventors, but by entities that might have been simultaneously describing
the same compounds as having no medical benefit. That was several years ago, but there are many examples.

Even some `open source` software has to be manually acquired and configured due to the licensing and other IP considerations.

What we were attempting to convey is that the techs `could be` at least during consideration stages, be reviewed for being licensed under, for example - not endorsing the concept here - a GPL type license. Where the work remains, in general, `free`, even if remixed. The subject is too extensive to go into here and now, but one could also draft their own licenses.

That is what happens when you have an owner. I don't see this project having an owner. Do you?

Currently, there is no clarified project. `Owner`? Depends. The contributors would certainly be the owners of their own efforts and work put into the project. The maintainer(s) of the project have responsibilities to make sure the releases are up to date, questions answered, requests responded to. Team memebers might be considered to perhaps `own` their volunteered time or `own` responsibilities to the team/project. Again, we reflect here on, for example, a currently popular operating system, for `PC`s, found its initial push with their terms of a specific licensing matter; where they were not bound to an exclusive distribution `deal`. Thus, they in turn licensed the same (already existing?) os, which, from what we could gather, was purchased for a nominal fee, to others around the world. Now the majority of the world might know that os by name. Probably much more to it, we were not there. The point being, if the entity that made that `deal` all those years ago could do it again, they might put in an `exclusive` clause somewhere in the text.

What we are posting is that not everybody is scrupulous, and to at least try to protect the integrity of the poject and its developers and persons that might use the tech - at least some consideration should be given to the area.

rives:
Something that you may want to give more consideration to is the overall architecture of your system.

--

Approach it however you like, but the system architecture becomes the foundation for all further component choices. If it isn't chosen up front, there can be a great deal of time and money expended in developing capabilities that won't be utilized in the end result.


Hi, rives.

Good posts and points therein.

Your thoughts on AOSP? As a platform? Already having ARM chip, sensors, networking, extensibility, and being based on, at least in part, POSIX?

We visited with relations recently. The lady of the house had damaged her device, and got an `upgrade` to a new device. Sold as a `phone`. A question came up relevant to some fact or the other. The man of the house said, "Honey, ask your phone!" Which she proceeded to do by simply voicing the phrase into the `phone`. The results were returned, and we then continued the discussion relevant to the `facts` returned. This occurred several more times. The man of the house also noted that not anyone could ask the phone the question - only the lady of the house - or the device would not respond, or repsond with an error message or alert. We will leave it at that, persons can draw their own conclusions as to the technology - but the technology is present - nearly ever present - and capable of all types of automated process, right out of the box.

Kind regards,
/SRGB/

 

[rives]

Hi, rives.

Your thoughts on AOSP? As a platform? Already having ARM chip, sensors, networking, extensibility, and being based on, at least in part, POSIX?

Kind regards,
/SRGB/

Never having used the equipment, I really don't have an opinion on it. However, I am very predisposed to using industrial gear. I've seen controls built on many different platforms, from S100 buss wire-wrap boards, to Apple II's, Timex-Sinclair 1000's, etc. In virtually every case, they died at the most inopportune moment.

PLC's are built for a completely different audience than consumer electronics. When you have a factory that costs 10's of thousands of dollars per hour (or minute!) to run, a custom-built controller running a critical process can get unbelievably expensive when it gives up. PLC's are frequently built to military specifications, will function in environmental extremes that would kill most electronics in a heartbeat, and just keep plugging along. I did my first machine automation with PLCs in 1981, and when that facility shut down over 20 years later, they were still in place and doing their job. Over a period of 30+ years, I've lost a power supply or two and a few boards - comm modules, a few I/O boards here and there, but never anything substantial after the infancy failure period.

This level of reliability is, to me, a fundamental requirement for our application. Murphy dictates that the damn controller will inevitably puke at the worst possible time, and usually after you have been lulled into a level of complacence that has caused you to misplace your documentation, be out of town, can't find the spares, etc, etc. If the gear is controlling life-critical processes, then the loss can easily go from being inconvenient to being devastating to your pocketbook and your commitments.

 

[Ttystikk]

Wow- I feel like I just went back to college and enrolled in a class on control systems and another on open source licensing issues... now I'm as confused as I was after the first day at school! Actually, it feels good. Thanks to all who choose to contribute!

As poorly stated above, I have a great interest in this and I have the facility to do some good testing, with multiple spaces, a fully integrated water chilling system, home HVAC, plans to replace the damaged driveway which could get geothermal lines very easily, more plans for aquaponics, a greenhouse and even installing fuel cells to see if their promise of greenhouse compatibility holds up. ALL the above is going forward sooner or later, and of course it needs to be effectively and efficiently controlled.

Whatever is developed, we can certainly start with one space, one RDWC to monitor and control and work our way up from there.

I'm no programmer, so my expertise there is more likely to be of the 'end consumer perspective' variety, lol

 

[VAtransplant]

Wow, this thread blew up I got confused reading it too hah.

What I will say is that it's Saturday and I've been having a great time screwing around with servos in the last hour. I got a bunch of rather high torque servos for $11 as well as a variable transformer with excellent reviews for under $70, but that hasn't arrived yet

My plan is to actually use one of the servos to control the position of the variac's knob. Obviously I'll know where the safe minimum and max are on the knob before I connect the servo, but this seemed like the easiest way to control my exhaust fan's speed safely depending on temperature. It doesn't appear as if my model is compatible with the standard speedster style controllers.

Anyway, in fiddling around with this and watching a few youtube videos, it seems like these servos are great for controlling duct dampers as well. Just gotta make sure you get one with enough torque to handle opening and closing with a certain amount of air flow pushing on it. I should have just gone that route and put a wye on my exhaust duct, controlling the position of the servo (and thus the damper) and sending air back inside until it gets too hot. Hopefully the variac thing works out.

I can't argue that arduino is the best thing out there, but for a small time grower who doesn't want to spend a ton of money getting it all set up - I think it's a great choice. I started typing up a giant step-by-step for 100% arduino/programming/electronics newbies to build a basic temperature/humidity monitor but it's on my work computer. I'll likely still throw it up in its own thread just to have it available for people searching in the future.

And for the record, if I were setting up anything short of a commercial warehouse style grow, I'd probably opt for the arduino. Even using a single one to control both your veg and flower rooms shouldn't be too difficult. It depends on what you want to control though I suppose. Temperature, humidity, pumps, lighting, and alerts? All cheap and easy stuff. If I had a second identical room, or third, or fourth... I'd just spend the $17 (free shipping!) for another Arduino UNO. If I had duct work set up distributing A/C to each room, I'd simply put a $4 damper on each outlet and control via servo. While arduinos aren't really industrial I would never question their reliability. Put it in a nice waterproof project box / pelican case, monitor your code closely for a little bit, and then let it do its thing.

 

[VAtransplant]

Wow- I feel like I just went back to college and enrolled in a class on control systems and another on open source licensing issues... now I'm as confused as I was after the first day at school! Actually, it feels good. Thanks to all who choose to contribute!

As poorly stated above, I have a great interest in this and I have the facility to do some good testing, with multiple spaces, a fully integrated water chilling system, home HVAC, plans to replace the damaged driveway which could get geothermal lines very easily, more plans for aquaponics, a greenhouse and even installing fuel cells to see if their promise of greenhouse compatibility holds up. ALL the above is going forward sooner or later, and of course it needs to be effectively and efficiently controlled.

Whatever is developed, we can certainly start with one space, one RDWC to monitor and control and work our way up from there.

I'm no programmer, so my expertise there is more likely to be of the 'end consumer perspective' variety, lol

If you want to start with a single room run, post up what equipment will be in it and I'll give you a price breakdown for the arduino route. Including how many pumps, fans, lights, other 120v stuff you want to operate and any extra features (like a way to alert you when your controller bucket is low for ex)

 

[hvac guy]

Thumbs up to Rives, I've installed 100s of PLCs and find them to be very reliable when it comes to critical functions.

 

[Ttystikk]

If you want to start with a single room run, post up what equipment will be in it and I'll give you a price breakdown for the arduino route. Including how many pumps, fans, lights, other 120v stuff you want to operate and any extra features (like a way to alert you when your controller bucket is low for ex)

Okay, just gonna take one and go for it. I'll keep it relatively simple.

One RDWC, currently with six tubs and one control bucket. One 1000gph circulation pump, to be running constantly if plants are in the system. Needs monitored for all the usual parameters;
Temp
pH
EC
Level

Room;
Temp
RH
CO²

Three 120V circulation fans, one for each silo.
Two 120V fans running small air handlers; one is an 8" brushless motor Maxfan, the other is a regular 8" muffin fan. Each of these blows down through an 8" Icebox water to air heat exchanger core, then down through rigid ducting to capture the distillate and return it to the RDWC.

Currently, I use a Sentinel CHHC-4 environmental controller to control room temp by operating the air handler fans and CO², by tank. I don't have a suitable heater, and it does not control RH; the chiller does indirectly, via its temperature setting. The chiller's capacity is shared with other spaces.

The room has a window I keep sealed, a closet with the doors removed and two doors. Not sure how relevant some of these details are, but I'm not leaving anything out. Unless I forget. Lol

I suspect this room of an air leak; the humidity is never as high as my other spaces and the CO² tank runs out faster. I'm investigating...

 

[Ttystikk]

The floor space of the room measures 10' 10" x 9' 3", while total air volume including closet and doorway alcove is 10'10" x 11'6" x 7'6" high = 935 ft³