Automatic plant watering project

Hi

Thanks to KellyW, naoki and others for asking and showing interest in my project. I hope to show all of you what is done, and what I plan to do in the future.

I work on a potted plant automatic watering device to help my plants to thrive. It is a hobby project. I made a simple microprocessor-controlled tool with minimum functionality and now I test two different ways to measure moisture to get timely, controllable watering. Also I intend to keep the cost down, so that multiple devices could be used without need to break the piggy bank.

The tool looks like this:

It is prototype, done in THT components. If ever comes to production, I will use SMD to miniaturize. Now it is 2" by 2".

It has the following functions:
One button calibration
Submerged, small pump with pipe to dispense water
Moisture probe
Long time battery operation (around one to three months), possible wall wart operation.
LED heart beat indicator

I researched two probes: resistive and capacitative. You can see both of them here. I am improving the capacitative probe right now, because it is not waterproof. I also plan to make a simple TDR probe.



The two red rods are carbon, as used in RC models. They are resistant to corrosion and allow to measure resistance of moist soil to some degree of accuracy. Water is in the pot below the plant or it can be in a container beside the pot. Water may go directly to the plant or to the pot base, whichever way the plant prefers. In the picture the pipe is hidden, because my Venus flytrap is full of water - not easy life for her with crazy amateur scientist around

The big flat probe is really a flat capacitor. Water in pot is a dielectric, so capacity changes can be measured to get some indication of water amount. It is better method, probe is isolated from soil, so no corrosion will occur.

The calibration button works in very simple way. When you press it, the device measures moisture, stores it in memory, then water start to run. When you let the button go it stops. The watering time is stored. From this point on, device works automatically. It will run the pump, when soil moisture drops to the point of the initial value.
Repotting or relocating electrodes requires recalibration. It is as simple as initial calibration. You need to power off the device, wait until watering is needed, switch it on and within one hour, press the button and calibrate the amount of water. Subsequent calibrations without switching off will change the amount of water, not base moisture.

That's the minimum version, it allows to go for vacation without worrying about your beloved plants.

I encountered some problems on the way. The first one is easy. The surrounding of the pot is quite aggressive environment, so good enclosure and moisture entry protection is needed. The second one is much more serious. I looked at methods of measuring soil moisture in professional solutions. They are rather expensive, and not needed in pot plant care. For example, salinity of soil and water quality influence must be taken into consideration. A big plantation can have different soils, with differing structure, chemical content and many other issues. The pot is the pot and soil or substrate structure is fairly uniform.

Still, even approximate moisture measurement is not easy. Resistive methods depend on water conductivity rather than soil saturation. Capacitance probes are better in this regard.

I think about weighting the pot as well, for orchids. If dry, the pot weight is much less than after watering. Also, there is a plethora of air humidity sensors, I have ordered one and I will make some research.

Also, I plan to make a more sophisticated device with real time, light, humidity and temperature measurements. Real time clock will allow to restrict watering around dawn time. However, the cost will be much higher.

I am looking forward for discussion and your opinions about my idea! What would you need to have a perfect plant care device?

Best regards,
Tom

 

Ray, you are absolutely right. I see it in my orchid substrate very well. However, it is one pot and the structure does not change too often if ever.
I wait for a good quality humidity and temperature sensor, maybe for an orchid it might work. I will put it in some enclosure on top of the pot, so that it will measure humidity of air rising from the substrate. It could be a very elegant way to measure, provided it would work
The goal is not an absolute, scientific tool, but specific, limited to one pot. Each pot will need calibration as described above.

 

Ray, one more thing came to me after reading you post. Thank you very much for this observation!
There are some sensors enclosed in gypsum, so they get some water from soil. They are quite expensive, and now I see that they cannot be used with orchids. There is no guarantee, that their water content will indicate substrate water content due to nonlinearity you mentioned.
So I scratch them off

 

Hi folks!

I am back and I have some news. Please take a moment to see.

I have a new prototype and this time it works very well. I tested it with my phal and drosera.

It looks like this:


Can't see much? It is buried deep and nearly invisible
The small pump is in the ceramic pot, there is pair of wires and pipe.

There is a detailed picture of a set.:


As you can see, I used flat capacitor to measure substrate capacitance. This way the electrodes have no electrical contact with soil.

I made some test using a small app I wrote to monitor the soil humidity and outside temperature. Here you have some screenshots.

Control panel:

Flower panel:


It seems that the humidity graph is quite accurate, the spikes are watering spikes. I noticed many interesting phenomena, for example during the day substrate humidity is rising even without watering. The reason is, bottom side of leaves expire some water to cool the plant. It more pronounced during warm and sunny days. It stops, when there is little water in the soil.

So, I decided to start production and I will post an indie campaign soon. I will keep you posted.

In the meantime, please let me know what do you think?

 

Ray, might be indeed. However I measure soil humidity, not air RH so, it would rather go down.
I plan to make some reference measurements in next two weeks, to get better grasp of the process. I hope it will make more clear picture.
One of my friends is soil specialist, he told me about transpiration:
https://en.m.wikipedia.org/wiki/Transpiration
He said that it might be possible explanation, but he had never researched this.