Are Commercial Growers Using Computer Technology in Their Grow Ops?

[Hackerman]

I know we have one commercial grower here (NorCalHal) and I think there are others, both legal and otherwise.

When I get out to Cali, I would like to get involved in the canna trade. Mostly for pure enjoyment but some extra cash might make it more fun

I am thinking what I might have to offer that isn't available on every street corner. My long suit is computers so I should probably look there.

I am wondering, do the larger commercial growing integrate CAD/CAM (Computer Aided Design / Computer Aided Marijuana)? LOL

I wouldn't think any grow op should be completely automated but I can think of a zillion things that I wold like to monitor and log. Everything from light sensors on each light to graph the diminishing output of light. This would allow you to change your bulbs at exactly the right time to maximize output per dollar and still utilize the bulb to it's fullest. Not a big deal to you and me but if you have 80 or 100 lights running, it would definitely save dollars in short time.

That's just one of a zillion things that I can think of.

The hardware to set up monitors like this is so cheap it's almost negligible in the overall cost. Writing the database interface and output will be a major task. I'm thinking that's where the money is. And, the initial setup might take some moderate effort. Still, this might be a fun project that could pan out and help some of the bigger growers keep tabs on changing conditions in the grow room.

As far as executing any automated action based on the monitoring data... maybe not a great idea. I think nute mixing, pH balance, trimming, etc might still be best handled by a person one on one with the plants. Maybe some kind of automated watering system, but that's about it.

Mainly this is for monitoring conditions and logging them. This allows statistical outlook at everything from air, light, water, etc, to growing habits in varying strains. That kind of information can save lots of dollars for business. Like I said, not a big deal for you and me but take everything you do and multiply it times 100 and all of a sudden, little things matter.

So, back to the original question.... does anyone know if the bigger growers are incorporating statistical technology into their grows and, maybe what they're doing.

Thanks


I am going to start playing with this on a tiny level. I have a Raspberry Pi that's pretty capable of doing a lot of data collection. I am going to put together a little wish list of things that I think might benefit a grow room. If you can think of anything, please feel free to add it to the list.

I think an optical guided guided laser cannon that targets and zaps spider mites might be cool.

Thanks again. I am really hoping to be out of this snow and cold before next Winter. So. Cali, here I come. I'm really looking forward to that stuff you guys have out there that we just don't have here...... like sunshine. LOL

 

[umbra]

I know that there is a system that will dispense nutes and ph it in real time, not fill up a reservoir and mix it. It is about $25K.

 

[NorCalHal]

Hi Hackerman

I see a lot of financial statistical programs being developed for cannabis, but not too much for the actual cultivation aspect. I like the idea.

For me, I would love to monitor PAR values and room conditions for each room. Keeping a record of those items would be a big help. A daily growth rate chart would be great also for various strains.
I see where this could go and I think it is a great idea.

As Umbra said, there is an automated feed system that I am going to implement that will dose the feed to set parameters before being fed to each plant site. The system I will be using is from Dosatron. Basically, you set PPM and PH and the feed schedule, and it does the rest. SUPER sick!

http://www.dosatronusa.com/

 

[Hackerman]

Thanks Hal. Nice to hear from you. I kind of let this thread drop for lack of interest. However, I am still playing with this project.

I have the sensors setup and running for temp, humidity, moisture, light intensity and light on/off. I'll have my camera next week for charting growth. I also have a proximity sensor but I'm not sure yet if I can use it to chart growth. Time lapse photography with the cam is going to be totally cool.

I have started the coding and I have the sensors reporting about every 3 seconds on each event. This part was actually pretty easy as a lot of the code has been done so most of it is editing, rather than raw coding. It will be a long road putting together the interface. I may see if I can team up with a couple other people to make things go a little quicker. Teams are always more productive. It's one of those, "the total is more than the sum of the parts" things.

Right now, I need to send the data locally on a computer. Since I am in an illegal state, I'm not real keen on shooting data from my grow room to an online database LOL so I'll probably run my own database for now. Eventually, I'll go with an online database because it has so many more advantages. A big one being that you can monitor your setup from anywhere. That will be a simple switch once I'm ready. Lots of stuff to do in between now and then.

The data that is collected is sent to 2 places. One is a dashboard that shows everything in real time. It shows in easy format of the viewers choice. See attached pic from another similar Pi project for a simple sample display of 2 events.

A graph charts whatever axis you choose. Time being one of them in most cases. However, phases of the moon might become important to you. Who knows? LOL Those could easily be incorporated.

The second place is the database where query can be made. Query possibilities are virtually endless dependent upon collected data. Data input is also virtually unlimited based on the sensors. But, you can also include things like moon phases, times of day for planting/harvesting and a million others. Combined with water/air/light/nutes/etc, you have a lot of data input.

I'm not sure what you mean by PAR. To me that's a type of lamp can we used for rock and roll lighting back in the day. LMAO I think it stood for Parabolic Aluminum Reflector. We just called them PAR cans. I think PAR 30 was our big thing.

So, if it means something in lighting, yeah, I'm in that to. The sensor I have will measure intensity so I can chart a bulbs useful life (I hope this sensor has that range). It can also chart off and on (obviously).

If PAR has to do with the air or water quality, I have not seen any sensors yet for air. I have seen pH and water sensors. Whatever doesn't exist.... can be built.

The good news about all this is that all the sensors and the main computer are still all well under $100. After all the research and coding is done, I could probably build one of these and put it in operation for a hundred bucks.

So far, mine is all geared toward human interaction based on the data. Few machines in this world work better than a well informed human. Machines are nice but I like a human touch.

Don't get me wrong, to have an automated watering system that dispenses water when the sensor says it's dry?.... Awesome. Simple stuff for simple machines.

But to chart growth patterns based on all those variables of air, water, light, moon phases, etc and combinations of variables. And, then to experiment and develop new growing patterns to program the machines on how to work..... well, that takes humans. Well informed humans. That's why Monsanto and every other big money grower still has chemists and biologists in laboratories gathering data and experimenting. LOL

Oops... sorry for the rant.LOL

I'll post updates as I get over major landmarks. Right now, I am really just poring over data and setting up the hardware. Pretty boring stuff if you're not into it. LOL

I welcome any comments on what someone might like to monitor in a grow room. Not sure how we can track insects or disease. But, when they come, we will know exactly what environment we had when we got it/them. And, if we find we get bugs/disease often and that environment is a variable every time... bingo, we just got smarter.

Anyway, thanks for the reply.

View attachment screenshot-raspberry-vivarium-controller.jpg

 

[Rosebud]

I bet there are a lot of folks interested, like me but had nothing to add. :woohoo:

 

[JustAnotherAntMarching]

I bet there are a lot of folks interested, like me but had nothing to add. :woohoo:

I'm in the boat with Rosie and we are just :48:

 

[Budlight]

Thanks Hal. Nice to hear from you. I kind of let this thread drop for lack of interest. However, I am still playing with this project.

I have the sensors setup and running for temp, humidity, moisture, light intensity and light on/off. I'll have my camera next week for charting growth. I also have a proximity sensor but I'm not sure yet if I can use it to chart growth. Time lapse photography with the cam is going to be totally cool.

I have started the coding and I have the sensors reporting about every 3 seconds on each event. This part was actually pretty easy as a lot of the code has been done so most of it is editing, rather than raw coding. It will be a long road putting together the interface. I may see if I can team up with a couple other people to make things go a little quicker. Teams are always more productive. It's one of those, "the total is more than the sum of the parts" things.

Right now, I need to send the data locally on a computer. Since I am in an illegal state, I'm not real keen on shooting data from my grow room to an online database LOL so I'll probably run my own database for now. Eventually, I'll go with an online database because it has so many more advantages. A big one being that you can monitor your setup from anywhere. That will be a simple switch once I'm ready. Lots of stuff to do in between now and then.

The data that is collected is sent to 2 places. One is a dashboard that shows everything in real time. It shows in easy format of the viewers choice. See attached pic from another similar Pi project for a simple sample display of 2 events.

A graph charts whatever axis you choose. Time being one of them in most cases. However, phases of the moon might become important to you. Who knows? LOL Those could easily be incorporated.

The second place is the database where query can be made. Query possibilities are virtually endless dependent upon collected data. Data input is also virtually unlimited based on the sensors. But, you can also include things like moon phases, times of day for planting/harvesting and a million others. Combined with water/air/light/nutes/etc, you have a lot of data input.

I'm not sure what you mean by PAR. To me that's a type of lamp can we used for rock and roll lighting back in the day. LMAO I think it stood for Parabolic Aluminum Reflector. We just called them PAR cans. I think PAR 30 was our big thing.

So, if it means something in lighting, yeah, I'm in that to. The sensor I have will measure intensity so I can chart a bulbs useful life (I hope this sensor has that range). It can also chart off and on (obviously).

If PAR has to do with the air or water quality, I have not seen any sensors yet for air. I have seen pH and water sensors. Whatever doesn't exist.... can be built.

The good news about all this is that all the sensors and the main computer are still all well under $100. After all the research and coding is done, I could probably build one of these and put it in operation for a hundred bucks.

So far, mine is all geared toward human interaction based on the data. Few machines in this world work better than a well informed human. Machines are nice but I like a human touch.

Don't get me wrong, to have an automated watering system that dispenses water when the sensor says it's dry?.... Awesome. Simple stuff for simple machines.

But to chart growth patterns based on all those variables of air, water, light, moon phases, etc and combinations of variables. And, then to experiment and develop new growing patterns to program the machines on how to work..... well, that takes humans. Well informed humans. That's why Monsanto and every other big money grower still has chemists and biologists in laboratories gathering data and experimenting. LOL

Oops... sorry for the rant.LOL

I'll post updates as I get over major landmarks. Right now, I am really just poring over data and setting up the hardware. Pretty boring stuff if you're not into it. LOL

I welcome any comments on what someone might like to monitor in a grow room. Not sure how we can track insects or disease. But, when they come, we will know exactly what environment we had when we got it/them. And, if we find we get bugs/disease often and that environment is a variable every time... bingo, we just got smarter.

Anyway, thanks for the reply.

This is awesome you have to keep us posted on how that goes I really wanted to set something like this up but have no clue how to do any kind of coding

 

[Hackerman]

I have been spending some time on this and it's coming along. It's been a while since I did any real coding so the return curve is still high. LOL It should flatten out quickly.

This Raspberry Pi is totally cool. Hardware has come a long way since I was into this type of thing. I have a full Apache SQL server running on on a "computer" the size of a beeper. There's also a lot of assist on the web. I am using Google for the API and it's pretty cool.

Essentially, I can monitor Temp, Humidity, CO2 content, Light Intensity (still looking for more things to monitor. Open to suggestions).

I can map any of those, so I can see, for example, temperature changes over the past 24 hours as the light cycles. Or, I can monitor light intensity over the past 6 months to a year to graph light intensity diminishing over the bulb life. Virtually everything can be mapped from the time the setup is placed into operation... to current date. Pretty neat.

It will take 6 months to put this all together but it looks like it's going to be a fun project.

I am definitely running into problems. The RPi only has digital inputs. All the best temp, humidity and CO2 sensors are analog. So, I will need to add an analog to digital convertor before I go too deep into the hardware. Right now, I am using all "hobby" level hardware to keep it cheap and fun. I can definitely see this developing into a more professional level at the sensor hardware level. Still, like I said, it's looking like a fun project.

Next is the moisture sensors in each pot. That should be pretty easy. I'm just doing 1 pot for prototype. Eventually, I might have some interaction like automatic watering or something.

I am always thinking outside the box and one cool thing about this kind of controller is that we can actually turn the light(s) off and on based on a diminishing cycle (like the real sun). I mean, the real sun doesn't really come up every day at the same time and go down at the same time to give a perfect 12 hour day / night (like we do with our grow rooms lights). I wonder how much effect a gradual change in light timing would have, both from a quality and/or quantity standpoint. Stuff like this and celestial influence are understudied IMO.

The time lapse pictures that it takes every day should really show changes like that. Actually I am taking pics twice a day. One picture an hour after lights on and one picture an hour before lights off. It's amazing just how different your plants look at different times of day. I guess it's just like us. I sure don't look my best first thing in the morning. LOL

OK, on a kind of related topic, here is something new that I am thinking/working on for commercial use........

electricity recovery.

I have only seen a few commercial grow rooms, and never one for real so I am really reaching here but.. how much light is lost? Let's say we placed solar panels to collect the "stray" light and recover some lost electricity.

Electrical costs for a commercial grower have to be a major expense. Even recovering 10 or 20% might be sweet.

Let's say we place the panels on the floor. We recover what little gets through the plants. Or, place them on the ceiling and get all the reflected light.

I'm not sure how the panels work so I don't know if they "steal" light but if they don't, they could be used as side panels on walls. I would have to look at reflectivity vs "absorption" but maybe even placing them between rows of plants.

Like I said, without ever really seeing a commercial op, I'm just stoned and reaching but, I have not read anything on it so I wonder if it's a new idea or an old one, proven useless.

Once I get the RasPi up and working, I can put some solar panels up on my walls and start playing with that.

I can't wait to get to a state where I can actually discuss this type of thing with people who have a real knowledge and interest. I can't say beans to people who live around here. Not a lot of commercial activity here, yet. At least none that advertise. LOL Viva la recreational cannabis trend.

 

[Hackerman]

I am getting ready to head out for a gig tonight but here are a couple quick shots (I guess this thread is my "Grow Journal" LOL)

The first is the little RPi with a breadboard that has the 3 sensors. I have a Temp/Hum gauge to help compare the numbers returned by the RPi with actual conditions.

From left to right on the breadboard is the Moisture Sensor, the Temperature/Humidity Sensor and on the far right, the Light Sensor. Next to the breadboard is the RPi with the camera sitting in front of it. I have been overheating the crap out of this thing so I have a super-fan cooling system. LOL

The HTML page is coming along. I have the temp and humidity gauges reporting properly. I have to make some adjustments on the charts. I will have temp and humidity and light (LUX) and CO2 gauges all across the top in a line reporting current conditions. Below that is everything in chart history charting the past 24 hours to past year. The Moisture sensor will only activate when soil is dry so no charting there. Only a WET/DRY option. The light sensor will map the entire year. The temp and humidity, probably the past 24 hours. At least, that's what I'm thinking right now. That's the nice thing about doing it yourself.... you can change.

Hope this isn't too boring.

View attachment pi-1.jpg

View attachment pi-2.jpg

View attachment pi-3.jpg

View attachment pi-4.jpg

 

[screwdriver]

What about a weight sensor? You would see a gradual weight increase and rate of increase. You could see when watered and how fast it lightens. That graph would be interesting to me.

 

[Hackerman]

That's an interesting view. I checked out the available sensors and most of them only sense weight, like a finger press or a footstep. None of them actually seem to weigh them. I'll check it out more. That might be a good time to tell when a plant needs watering. Combine that with the moisture sensor and you would have a pretty failsafe setup.

I did add a few new features. A window next to the gauges. Right now I have it so it simply takes a picture in the morning and displays it in the frame. However, I think I'll change that to live streaming and enlarge the window a little bit.

I added LUX and I actually found a formula for converting ohms resistance to LUX and LUMENS. Thanks to Zorro for this.

I'll try to proceed stp-by-step.
We convert the log-log graph into a linear one by this change of variables:

x = log10(I[lux])
y = log10(R[kΩ])

Then we have the same graph but the abscissa axis is linear (ranging from -1 to 5) and the y too (ranging from -1 to 3).
It is a straight line whose equation is

(y-2)/x = (y+1)/(x-4)

It simplifies to

x = -4/3 * (y+1) + 4

Then, reaplacing and grouping terms, etc...:

log10(I[lux]) = -4/3 * ( log10(R[kΩ])+1 ) + log10(10000)

log10(I[lux]) = log10(R[kΩ]*10)^(-4/3) + log10(10000)

log10(I[lux]) = log10 [ (R[kΩ]*10)^(-4/3) * 10000 ]

I[lux]) = (R[kΩ]*10)^(-4/3) * (10000)

I[lux] = 10000 / (R[kΩ]*10)^(4/3)

Regards

Z

I just got my second RPi so I'll be putting one in the grow room today and I'll start tweaking the code based on real time events. Pretty neat stuff.

View attachment chart.jpg

View attachment pi-5.jpg

View attachment pi-6.jpg

 

[Hackerman]

This is a pretty lame example but I'll toss it up here just to see how it looks.

I bought the camera designed for low light (duh) so I am waiting for the proper one to arrive from China. Probably a couple weeks I'm also waiting for the case to mount everything permanently so the pics are from exactly the same perspective. You can see the camera moved. This is only a couple weeks worth of shots. I should have the bugs worked out in time to get a 'beginning to end' animation for the entire flowering cycle on my next crop. I should also be able to filter out the bad coloring that is so common in the flowering room. Lots of bugs yet to work out but still, pretty neat.

View attachment Animation2.gif

 

[Hackerman]

I spent some time this week on the HTML interface. I combined the temp and humidity charts since they are so related (as you can see). If you mouse over a location on the chart is shows a flyover of the actual date/time and temp (or humidity) reading.

The LUX is showing 7 days but I think I should do that in weeks so you could see the degradation of the bulb. I moved the light sensor away a little bit each day to simulate what a failing bulb might look like. Ideally, each light will have a sensor and it will be attached directly to the light's lens. The sensor is only about the size of a bobby pin so it would clip nicely right along the edge somewhere.

The picture on the top right will be a live stream of the grow area and clicking it will take you to a new page where you can pan, tilt and zoom the IP cam that's in the room. The IP cam is all set up and working. All I need to do is write the interface.

I couldn't decide on a background so I just took a picture with the cam every 15 minutes. It's tiled now but I'll figure out how to stretch it. Might be cool if I can make the background a live stream and have the plants waving in the breeze in the background. LOL

Well, that's about it for now. I want to clean up some of the code and set up the moisture sensors for the plants next. Then, I'll put one in full operation in my veg tent.

Pretty fun stuff.

View attachment screenshot4-13.jpg

 

[Hackerman]

Well, this proves that even a blind squirrel finds a nut now and then....

I messed up and bought the wrong camera chip. I bought another one and I am using the wrong one while I wait for the new one to arrive. The one I bought is an infrared version (I was wondering why I could still see my plants with the lights off. DUH).

So, I was doing some research on how this might work in a room with a high brightness value like the grow room (since it was designed for low and no light applications) and I came upon this...

https://www.raspberrypi.org/blog/whats-that-blue-thing-doing-here/

It's an article about the camera and how biologists are using it with a blue filter to view and photograph photosynthesis in plants. Duh again. I wondered why the camera I bought came with a separate blue filter. LMAO

So, as it turns out, I might be able to use this camera in the grow room for some really fun stuff. Like, maybe time shooting to determine when the plants are at their top of photosynthesis. I love this stuff. Even a blind squirrel..... LOL

Here's a little teaser to the article....

If you’ve been reading our posts about why we developed an infrared camera board keenly, you’ll have noticed that we mentioned a lot of interest from botanists, who use infrared photography to work on the health of trees. We started to read up about the work, found it absolutely fascinating, and thought you’d like to get in on it too.

A short biology lesson follows.

Photosynthesis involves chlorophyll absorbing light and using the energy to drive a charge separation process which ultimately generates oxygen and carbohydrate. Here’s a nice picture of the absorption spectrum of two sorts of chlorophyll, swiped from Wikipedia:

[Chlorofilab]

Notice that both kinds of chlorophyll absorb blue and red light, but not green or infrared.

So: why are trees green? The graph above shows you that it’s because green is what’s left once the chlorophyll has grabbed all the long wavelength (red) and short wavelength (blue) light.

Let’s say you’re a biologist, and you want to measure how much photosynthesis is going on. One way to do this would be to look for greenness, but it turns out an even better method is to look for infrared and not blue – this is what the filter lets us do. Bright areas in a picture filtered like this mean that lots of photosynthesis is happening in those spots.

There’s a long history of doing this stuff from space.......

 

[Hackerman]

Got a new cam for my grow room.

I must say, it takes a pretty great picture in the dark. I was a little worried about the IR light but everything I read set me at ease. Apparently the IR range is outside the scope of the cannabis plant.

View attachment nightvision.jpg

 

[Rosebud]

Wow that is very cool hackerman. This way we can see if they are behaving themselves in there.
Do you think an electric toothbrush charging with a green light will bother my plants? You know, just a blinking light?

 

[Hackerman]

I am paranoid about light leaks so I put black tape over all my little LEDs in my grow room.

Granted, with all the toys I have in my grow room, there would be about 20 LED lights so...... LOL

I'm sure one toothbrush is OK.

I sit in my room after lights out and look around. If I see anything, I'm taping, sealing, caulking. LOL

 

[Hackerman]

Well, the new cam for the Raspberry Pi came in today. Pretty sweet. I am a little disappointed in the clarity and it's a fixed focus lens so I will have to play with adapters to get it just right but for something that's the size of a dime, it takes a pretty good pic.

I was definitely impressed that it's one of the few cameras that I am able to get a "green" shot under HPS lights.

By the next grow, I'll be ready to do time lapse.

There's a total of 5 cams in my grow rooms, now. LOL

View attachment image1.jpg

 

[Budlight]

I spent some time this week on the HTML interface. I combined the temp and humidity charts since they are so related (as you can see). If you mouse over a location on the chart is shows a flyover of the actual date/time and temp (or humidity) reading.

The LUX is showing 7 days but I think I should do that in weeks so you could see the degradation of the bulb. I moved the light sensor away a little bit each day to simulate what a failing bulb might look like. Ideally, each light will have a sensor and it will be attached directly to the light's lens. The sensor is only about the size of a bobby pin so it would clip nicely right along the edge somewhere.

The picture on the top right will be a live stream of the grow area and clicking it will take you to a new page where you can pan, tilt and zoom the IP cam that's in the room. The IP cam is all set up and working. All I need to do is write the interface.

I couldn't decide on a background so I just took a picture with the cam every 15 minutes. It's tiled now but I'll figure out how to stretch it. Might be cool if I can make the background a live stream and have the plants waving in the breeze in the background. LOL

Well, that's about it for now. I want to clean up some of the code and set up the moisture sensors for the plants next. Then, I'll put one in full operation in my veg tent.

Pretty fun stuff.

If I was to pay you do you think you could set something like this up to monitor and control PH

 

[sopappy]

I am paranoid about light leaks so I put black tape over all my little LEDs in my grow room.

Granted, with all the toys I have in my grow room, there would be about 20 LED lights so...... LOL

I'm sure one toothbrush is OK.

I sit in my room after lights out and look around. If I see anything, I'm taping, sealing, caulking. LOL

I can sit in my flower room (after eyes adjust)
and can't see a white card 6 inches in front of my face
and those red leds are hermie helpers, might get away with green but I wouldn't

I thought my grow was automated, holy crap, your moniker is pretty apt alright, very nice