Some of you have noticed that the lobby plants get neglected quite often and we can correct that by designing and installing a watering system that monitors each plant and supplies water when needed. We are currently looking for volunteers with all levels of experience from none to expert to design, build and install the system. If you are interested please fill out this FORM and we can get started.
1 Like
I built this last year for my office plant. I call him Lilā Pumpy. Works pretty well, plant is still alive.
EDIT: This can run standalone with Tasmota, but I connect it to OpenHAB to run on a schedule.
4 Likes
An exciting electronics project to be sure, and I think itās always worth doing whatever will be fun and a learning opportunity, but Iām wondering if it has major benefits over something like these Plant watering globes ?
3 Likes
Fan of the Wemos D1 / ESP8266 here. What is that web UI and how is the MCU communicating with it?
Does that device have a capacitive sensor or are you adding water without considering soil dryness?
1 Like
@Iammikecohen the tasmota firmware has a lightweight UI (pump01.lan ui), but I really just use that to initially configure the module: set pin configurations, wifi, and MQTT settings. Wemos (tasmota) pub/subs to MQTT. I have OpenHab (Nik Plant Watering Schedule ui) + mosquitto (mqtt broker) + grafana running on a raspberry pi.
I have a capacitive soil sensor connected, but I find it very unreliable. I switched to a weekly watering schedule.
2 Likes
2 Likes
Iām really late, but David mentioned this project to me today, so Iād like to bump the thread and see what I can do to help get the ball rolling. 
Who else is still interested?
Some of the plants are doing alright, but others are barely holding on! 
I have experience with microcontrollers, soldering, web development, and Python, but controlling water pumps is not something Iāve never looked into.
Iām sure there are a lot of ways to approach that from an engineering standpoint. Does anyone have a vision for how the plumbing for multiple plants with differing watering schedules might be controlled?
Multiple peristaltic pumps seems inefficient/overkillā¦ Maybe a reservoir above the plants thatās scheduled to open/close open and close valves attached to tubes that run to each plant?
I donāt know much about horticulture, but Iām confident I can figure out how to plan a āper-plantā watering schedule and write the code to control the hardware, itās just a matter of deciding on the right hardware and working backward from there.
Looking forward to hearing otherās thoughts about the best way to automate the plant watering!
3 Likes
Iāve been going through Arduino training programs during the week. Lots to learn!
I believe there are 7-10 plants, all hanging in the front window. David and I have batted ideas around for a while now. Some of these are āwould be nice, but not necessaryā. Hereās a synopsis of what we discussed:
- Separate water flows for each plant
- Moisture meter for each plant
- Means of stopping water pump before reservoir runs dry.
- LCD screen on wall for controlling settings and possibly seeing reports
- Light Meter
- Alarm options for too dry - too wet - too dark, etc.
- WiFi to send reports and alarms
- Keeping in mind that this is a highly visible project as it sits in front window. We would like it to be a show piece. Maybe something like status lights in each plant? Blinking light when something is needed? LCD Panel with constantly updating info and graphics?
First consideration:
How to control water flow to separate plants so each plant gets an individualized amount of water?
Controlling water to individual plants sounds like a later phase feature to me. If you want to see progress on this project I suggest you take a more agile approach. Donāt design the whole thing upfront before you start building. Get something out there and build on it.
I suggest we get the hub (which you labeled arduino) online and wire up one plant to start gathering data. @nikcollins ās stack is what I would use.
1 Like
I have to say at first I found this frustrating but I now genuinely love how complex this has become. 
1 Like
While home with a spectrum outage, I did some parts research. I may or may not have a window full of plants I want to do this for at home as well, so Iāve been idly thinking about it. 
Weād need these parts:
-
Control. Microcontroller clearly, with enough data pins for about 4 per plant - on/off or rate out, moisture level in, maybe feedback on water amount of possible. 1 more forā¦ RGBLED status light. Maybe soil pH or other data? Plus extra global pins to drive indicators, read water reservoir status, etc.
-
Reservoir. Could be a camping water jug or something, 1-2 gal would be fine. Preferably dark to avoid algae growth. BUT what if the reservoir was pretty - water in motion grows less algae, and a fountain running low is a good indicator. Plus it could be the overflow catch solution if large enough? No idea how weād do this!
-
Distribution. Some kind of 1-many splitter manifold, maybe expandable? Tubing - likely flexible fish tank type, 1/8" ID would be cheap, easy to design for and print cable management items.
-
Valve. DC electronic control valve, at really moderate pressure. Not ālow pressureā - that is specialist stuff apparently, and not high pressure either. We could possibly use something meant for pneumatics even, or just squeeze a flexible tube. Are pneumatic controlled water valves cheaper?
Misc notes:
McMaster wants $82 at least for any electronic solenoid valve. Their selection is way over spec for this application. They do however have shiny aluminum distribution manifolds, OR we could CNC some nice designs on the block facing in and out, and drill +thread manifold holes ourselves?
This place has one for 12V for only $36. EC-2-12-H I really donāt think this should be that expensive though - why is it? More Amazon and AliExpress delving to do here.
The Sparkfun soil sensor seems fine, but even their text alludes to reliability issues. I would anticipate 3d printing a casing for it out of transparent petg to allow water-protected light sensing.
I unfortunately have not been able to think of a way to reduce the per-pot BOM, even by spending more money on the system-wide bom.
Internet is back on, do back to work for me. Thought Iād share findings while I had them!
Well, I got about 5m of slow Internet before it went down again, so Iām back in the phone.
I came across this project: https://opensprinkler.com/faq/
Which led me to this supplier and product: 1/8" Nylon Electric Solenoid Valve 12V DC Normally Closed NBR
12vDC, 1/8" id tubing, and only $13.49/ea. Much better!
Adafruit sells a solenoid for $7, Amazon has a similar breakout two for $10
Speaking of Adafruit, this looks pretty handy https://www.adafruit.com/product/4026
The capacitive sensor does look good! I2C is a good bonus. I think that means it can expand with more plants?
I had initially discounted the adafruit solenoid because itās a 1/2" diameter. I think we may have trouble getting good control if we have to adapt it down to lower volumes through 1/8 or 1/4 tubing.
I just bought two of these. I think I have everything else I need to build a system like this for my planter garden.
Hereās a clever system to deliver water with minimal electric valves once the amount of water per period was determined.
Could also modify the pinch valves to actuate with RC servos instead of using more costly valves.
Another way to meter water that might make for cool-looking gadgetry is with a peristaltic pump. Make a big one and use valves for output routing, or make a little one for each plant. They could be laser cut or CNCed with a stepper motor to turn the rotor. Degrees to milliliters would be easy to calibrate.
1 Like