9.2.4 A crash course in grow room automation

By Pete Tewes, Aka GrowGeek

IoT Integration Engineer

Pt3 Technical, LLC.

***Full disclosure***  While I do happily work for a controls manufacture and have my own Engineering business that has a preference of the control line we use (it’s my personal favorite), the basic information provided in this article is brand agnostic.  I might reference specific manufactures components, but at the end of the day, at the basic level…  A relay is a relay, a sensor is a sensor, and a controller is just a controller.  Only difference in the controller is how smart the person programming it was, or what limitations they might have had!  😉  Differences in sensors are how accurate and if the calibration needs to be traceable for regulatory purposes.  If you get your jollies working with brand X controller and brand Z CO2 sensor, do not let me ruffle your feathers.  I don’t care to get in a pissing match on why one brand is superior to the next.  That is not what this article is about.  Please understand that any quality commercial building automation controller will be able to read and control all the same industry standard sensors and devices.  The maker boards from China?  YMMV.  Good luck Chuck!  😉 

We all know that plants do well when given optimal growing conditions.  How we get to those optimal conditions has many paths, but in the end, (hopefully) they all lead to a bumper crop!  I’m hoping my ramblings will help you build an understanding of what it takes to automate something like a grow room.   I am also hoping to help explain some of the basics of general automation and demystify all the inputs and outputs you might run across while contemplating your own automated grow.  I’m going to attempt to remove some of the smoke and mirrors that some people associate with controls and automation too.  I have been doing controls for as long as I can remember, and if I can do it…  You can too! 

Come along for a fun filled journey as I lift the hood of my current automated grow room project and share with you some of the neat things you can do once you have a few inputs, some outputs and a processor in the middle of them!  This system that I cover is probably a bit overkill for the closet grower but can easily control a small to large scale facility.  The limit really is your desire and pocketbook.

For the small scale indoor recreational grower that has no desire to have anything other than super basic control, the module type systems like Titan Controls (no affiliation) are really where it’s at.  Purpose built boxes, plug both ends in, hang up the sensor part (if external), crank some knobs or mash some settings buttons and you are good to go.  They are geared towards the consumer as a plug and play system. 

Good results can be had with some basic temperature, humidity, and CO2 control.  I am not going to argue that at all.  But If this is the extent of your desire for controls knowledge or grow room control please STOP HERE! 

There is nothing wrong with this kind of approach.  You will get “results”!  How good or bad the results are depend on more than just a few manual dials set.  I was this grower for many years when I first started growing.  I had a A/C unit set to 78 DegF, an oil heater set to about 68 DegF, my dehumidifier set to 45%, and a big ass mechanical contact timer for the lights and fans.  I also thought I was a big-time grower with my overpriced CO2 controlled relay that was attempting to keep the CO2 levels up.  How accurate that CO2 sensor was is a thing of strong opinion with me.  But the setup worked.  I was harvesting ok plants, but nothing spectacular.  With no ability to monitor my grow room remotely, I pretty much felt like I could not leave my grow.  In fear of something going wrong and me not being there to notice it.

Some of the more maker type growers have turned to off the shelf/internet automation devices and platforms like X10, zigbee, Arduino and raspberry pi to give them a bit more flexibility when it comes to component selection and gives them the ability to build a control program based on how they want the grow room to operate at a much more reasonable cost vs. most commercially available systems.  These require quite a bit more knowledge in electronics and programming than the plug and play boxes but do have their place.  Sensors are cheap, but accuracy might suffer a bit when mass produced in China and sold bare board.  Do not get me wrong, I’ve seen some amazing opensource projects developed on the Pi.  Lots of wiring diagrams are available with a simple google search.  Along with all the parts lists and sample code.  However, most of these projects will never be a commercially viable solution, nor would I want to trust my personal grow room to them.  But their underlying operation is the same as the basic “plug and play” and commercial control systems.  All have the same inputs, outputs and logic in the middle.

When I look at any grow room automation project, I first ask the client what exactly they are looking to get out of their “automated grow room”.  Because for a lot of people that means many different things.  Most everyone wants basic temperature, humidity and CO2 control.  Everyone wants control over lighting and wants to know when things are not as they should be.  All growers want to have a better awareness of what is going on in real time so they can make appropriate changes to better their yields and save money.  With the coming of the smartphones and IoT technology, more growers are wanting the ability to securely monitor their grow rooms and make changes remotely.  They want to be notified if anything is out of parameter or is going awry right on their phone.

When you start talking to the hydro crowd, they get excited when you talk about having the ability to monitor and control things like the pH and temperature of their nutrient solution.  When you are talking to the soil crowd, they get excited about the thoughts of monitoring their soil temperature or soil moisture.  The thought of having the ability to control the pumps and valves of each grower’s respective system just how they want brings joy to their hearts.  

There are a myriad of other possible data points that a grower might have a desire to monitor.  For example, monitoring wastewater runoff has become more of a focus in some jurisdictions.  Having the ability to accurately monitor, trend and alarm that runoff is becoming more crucial.  The ability to produce a monthly report is a lot easier to have a machine do, than have a person manually record and keep track of every day.  The controller is less likely to call in sick to work or forget to take that sample last Thursday.

Once the specific requirements are noted, then it is time to start looking at sensors.  Like the robot Johnny 5 says, we NEED INPUT!  We cannot control what we cannot see.  Just remember that.  If you want to control something, we need to be able to ACCURATELY MONITOR it with the controller way before we decide HOW to control it.  How to control it comes later when we start talking outputs.

Most commercially available sensors come in a wide variety of ranges and output types.  Appropriately choosing your sensors makes all the difference when it comes to good control and reliable operation.  They make sensors that monitor everything and can range from a few dollars to several thousand depending on what you want to monitor and how accurate and repeatable you need the readings to be.   

The basic 10k type 3 thermistor for temperature monitoring works fantastic in a grow room application.  +/- 2 Degree accuracy is fine.  I’ve also had good success using thermistors to monitor soil and hydroponic solution temperatures.  No need to get fancy, just make sure it’s waterproof and corrosion proof if you decide to go throwing it into a nutrient reservoir or something similar! 

If your application requires more accuracy than a thermistor, you can take your pick from many different flavors of temperature sensors.  Endress Hauser makes some pretty kick ass temperature (amongst other) sensors.  If you need that level of sensing accuracy, they might be worth a gander.  One could do a whole article on different types of temperature sensors and the different applications for each.  Hmm, might do that sometime.  Any rate….

Humidity sensors are about the easiest to work with as most are 0-100% humidity to 0-10 volts output.  Pretty easy to scale that one.  5.5 volts equals 55% humidity, 4.3 volts is 43% humidity!  See how easy control automation is!?! 😊  For the most part all the commercially available humidity sensors are the same quality and response.  +/- 2% full scale accuracy is good enough for most applications. 

Again, if you need tighter tolerances they can be had.  Cost goes up when you start talking calibration certified sensors, but if your requirements deem it necessary, it can be had.  Visalia are my personal favorite humidity sensors if the application and pocketbook call for it.  Not affiliated with the company, they just make really kick ass accurate products that I have personally used in many different control applications with great success.

CO2 sensors are where I personally say spend the extra dough on quality.  Most of them are all 0-2000ppm and can do all the standard 0-10 volt or 4-20ma signals any commercially available controller can read.  For high end, standalone dead accurate CO2 sensors, I have not found anything that can beat a Visalia.  Do most people need that detail of accuracy, no!  But know if needed, it and many more like it are out there.  Just please do not buy the eBay special for $30.00 and expect it to have the same response and accuracy as a $200.00 commercial sensor.

In recent years all the major control manufactures have come out with combination sensors that do temperature, humidity and CO2 all in one sensor.  This makes it easy for the installer as there is only one physical device to install.  However, it might not be the most ideal place for all the sensor locations, so you need to take that into consideration as you go.  Some manufactures even allow for multiple Temp/Humidity/CO2 sensors to be attached to one controller without using a single physical input of the controller…  Pretty snazzy feature when you want to monitor multiple points of a grow room and want to save all the controller inputs for other devices.

Current sensors are devices that can monitor if a device has an electrical load or not.  They have a small contact in them that closes when there is current and opens when there is not.  This is what the automation controller is monitoring.  These are good for when you want to make sure the fan you just turned on really is on!  Or if the light fails, you want the ability to detect that the ballast is not pulling any power anymore.  Just because we tell a device to run, does not mean it’s going to follow our direction!  Current sensors give us the ability to keep an eye on the status of things of that nature.

Once we have all the basic inputs roughly mapped out, it is time to start looking at the controlled equipment.  What do we need to control and how are we going to interface with it is the main question to ask when looking into automating a grow room.  The higher amp draw equipment like the lights and larger fans and pumps will require electrical contactors to interface the controller to.  The lower draw stuff can use panel relays. I use panel relays for all low voltage control circuits and most line voltage (Under 15A) loads.  You can pretty much count on anything that is an “on/off” type control will be using a relay or contactor for interface.  The controller will supply voltage to the relay (or contactors) coil.  That will close the contacts and power the device.  When the controller stops sending voltage to the relay, it will open the contacts and the device will no longer have power.  Lights, Heating enable, Cooling enable, Humid/Dehumid control, CO2 solenoids, 2 position irrigation valves and dampers, non-variable speed pumps will all fall into this category of control.

Once we know what inputs we have available to us and what equipment we will need to control, we then start developing our own control strategies for each aspect of our grow room. 

  • The lighting system will always follow some form of schedule, but could have a strategy (for instance) that stages 4 banks of lights on and then back off to simulate sunrise and sunset.
  • Heating and cooling can be as simple as a basic thermostat control strategy. When the temperature drops below the heating setpoint, turn the heater on.  When the temperature is above setpoint, turn the heater off.  Same goes for cooling.  Above cooling setpoint, crank the AC.  Below cooling setpoint, turn the AC off. Or it can get very detailed about looking at multiple sensors and making decisions based off time of day, humidity levels and any other variable you might want to reference.  Some growers find benefit in having the ability to do controlled temperature swings during the day to more simulate outdoor conditions.
  • Humidification and Dehumidification control strategies (for the most part) are basic humidistat control. Again, some growers have found benefit in having the ability to adjust humidity levels based off other variables. 
  • CO2 control is where everyone has a different take on the subject. Most of the cheap sensor enabled outlets that are being sold today are just a CO2’astat (it sounded good to me).  They look at the setpoint and if the CO2 level is below the setpoint, they energize the outlet and that powers your solenoid.  When the CO2 level goes above the setpoint, the outlet is deenergized and the solenoid valve closes.  Most of the commercially installed systems are like this too…  When you are given the opportunity to have multiple CO2 monitoring points and control over fans and such the possibilities become limitless.  Averaging sensors or looking for the lowest value.  How best to control CO2 levels with an automation controller is something (again) that could take up an entire article.  It’s also another topic that I love to engage other growers on.
  • Hydroponic systems will have their detailed control strategies as will soil systems. Some could be as simple as opening the watering valve twice a day for 15 minutes.  Or it could be as detailed as batch making nutrient solution, with chemical feed pumps and various solution sensors.  Monitoring and controlling a chiller could be a requirement if nutrient solution temperatures are to be controlled.

Once the control strategies are in place and the system has been commissioned to the point that you are confident in its operation and control, we can the start to look at ways to remotely monitor the system.  There are many different IoT enabled monitoring solutions out on the market today.  Utilizing commercial building automation controllers allows us to utilize all the industry standard communication protocols and allows us to interface directly with a large volume of mechanical equipment and supervisor type devices via communication lines and not physical i/o from the controller.  More and more equipment manufacturers are providing BACNet or Modbus communication interfaces.  Making our job to control and integrate with them even easier.

I personally use and install BACNet general purpose controllers and recommend a cloud-based service with a secure local gateway to get all of my clients grow room data into the cloud.  Yes, I work for the manufacturer.  No I don’t get a commission on product sales.

GG's Grow Room Panel

This is a picture of my grow room panel that I use (all current panels I install are identical to this).  It gives me a great universal platform to interface anything BACNet to the cloud, and being a general-purpose controller, I can customize the inputs and outputs as I see fit.  The ability to add I/O modules as I go is as simple as addressing a controller and connecting it to the network. 

I like the ease of being able to use a secure cloud-based system to monitor other devices that are onsite as well as my automation controllers.  Commercial power metering and security system monitoring are becoming more commonplace now.  Having all the trending securely offsite gives you limitless opportunities to crunch the data however you see fit. 

Data analytics for grow rooms is a rabbit hole I would love to jump down with anyone who wants to talk about it.  I am always fascinated by all the different information people want to extract from their grow rooms and how they want to use that data.  Here are some examples of some dashboards and trends from a recently installed control system.

Dashboards

 

Dashboards made simple.  Quick and accurate info at your fingertips.  Oooh…  Magic…..

Trending graphs for at a glance historical review.

Trending Graphs

Export any trend data to CSV for use elsewhere.  All securely stored in the cloud until you need to retrieve them.

Stored Data

As you can see, there are many ways to approach automating a grow room.  I hope that I have demystified some of the automation smoke and mirrors for you, and if you by chance happen to need some assistance in figuring out your own grow room, feel free to reach out anytime.  Automation is one of my many passions and I love helping others be successful with their own systems.

Until next time,

 

GG

 


Print  

Comments powered by CComment