[Daniyal]’s goal is to build an automated garden that allows him to grow plants in any environment he chooses. He’s got a good start with this rig, which is controlled by a Pi Zero connected via serial to an Arduino Mega clone, which in turn controls a bank of relays and sensors.
Monitoring the environment is a temperature and humidity sensor as well as a series of six soil moisture sensor spikes. The relays control the water pump(s?) and lights, allowing [Daniyal] to maintain specific conditions depending on what he’s growing.
[Daniyal] has ambitious goals for the project. The Pi has a camera on it, and he hopes to not only maintain the greenhouse from the Internet, but also figure out how to monitor plant growth automatically, so that the Pi can measure plant growth and adjust the conditions without his input.
We’ve covered a lot of very cool horticulture projects here on HaD, including radio-connected soil sensors, using G-cal to create an internet of lawns, and the Garden of Eden watering kit.
The cress.space (<-that's a website) project tries to achieve the same.
If you’re using Sparkfun soil sensors, they have been known to fail if left in the soil continuously. Make sure the tips are coated in gold and you power them only as needed to check readings. Otherwise, great project!
This kind of sensor is not even callable “humidity soil sensor”. It is merely resistivity/conductance sensor where humidity is a factor, but not the only one.
The only real way to measure water content is by carefully taking a sample, precisely weighting it, drying it in controlled condition and precisely weighting it… So unpractical and destructive measure.
So yes it’s true, it’s a conductance sensor… and it’s what professional use as an “humidity soil sensor” because it’s the only practical way and precision is good enough for growing crop… (Well, one of the two practical way: the other is by measuring it’s the dielectric constance, aka capacity sensor…)
I’ve been given a lot of recommendations to get capactive sensors instead of the resistive ones, which I will do in the future. At the moment to extend the life of resistive sensors I do not keep them on constantly. I have hooked up the vcc pin of the sensor to digital pin, this way I have to apply power to the pin to get a sensor reading. By doing this I can turn on the sensor, take a sensor reading, and then turn if back off for say, 15 minutes, and then repeat. IN this way I only have to keep the sensor on for about 3 milliseconds to take data, therefor increasing the lifetime of the sensor by a lot. But I still have plans to use capacitive sensors.
That is a issue I have discussed though the solution was not coating the electrodes with gold, nice idea. I do have the sensors VCC pins hooked to a digital pin. This way I can turn the sensor on only when I need to take a reading making the process of turning the sensor on, taking a reading, and turning back off take about 3 milliseconds making the sensors last a lot longer. Later I plan on investing in some capacitive soil moisture sensors since they do not seem to corrode at all.
That’s an interesting idea, I never thought of coating the sensors in gold. I’ll look into it. At the moment I do have the sensors set up so that their VCC pins are powered by a digital pin on the arduino. This way I can turn them ON only when I need to take a reading, as you suggested. I take a reading about ever 10 minutes (still experimenting with how far apart data points should be and still provide meaningful data) by turning the sensor ON, taking an analogue reading, and then turning back OFF, the ON time of the sensor measuring in to about 3 milliseconds. Hopefully this will increase the life of the sensors by a lot.
Garden projects seem to be popular with Arduino or Raspberry and though I have seen many, I find each and every nwe one interestin
It’s because openAG
my garden needs automatic dead plant disposal.
How about this? http://hackaday.com/2017/05/11/just-in-time-for-summer-a-remote-controlled-snowblower/
Lower it a bit and it should dispose of all dead plants. And it could even be considered a hack in two ways:
1. you use the snowblower in a way it was never intended to be used
2. you circumvent the need for detection of dead plants by making sure all plants passing through the device are dead, therefore only disposing of dead plants (as per your requirements)
How about programming the FarmBot to do it? https://www.youtube.com/watch?v=8r0CiLBM1o8
Flipping the terrarium upside down might make moving plants or replanting easier to do. Maybe place them in a shallow tray so you don’t have to worry about water running out. A thin intermediate coupler could also house fans and vents.
I gotta wonder why the electronics are in their own sealed off section rather than outside the grow box. Your sacrificing grow area and asking for humidity to mess with them if your gasket gets a leak.
What is the dirt for? Commercial green houses don’t use dirt. They have bags and growing medium, but it isn’t soil, probably for consistency and process control, not to mention pathogen elimination.
https://www.youtube.com/watch?v=k6lptZJ-kE4
I am working on a similar project. The intent is to automate growing plants on window sills. The objective is making growing lettuce, kale and spinach easy to grow on window sills in the Canadian Arctic.
Here’s the link to the project page: https://hackaday.io/project/16298-agi-esp8266