Motivation - Meet Bob
The layout of a space really affects how you feel in it, and I am always looking for ways to improve the feel of my workspace. So I decided to get a plant that could live in the windowless lab I spend the majority of my time in. His name is Bob.
Almost immediately I realized that between my lack of plant care experience and my moderate travel schedule, the plant would need some help getting watered.
My sister and her boyfriend thoughtfully bought me a Raspberry Pi for christmas, and now I felt like I had the perfect project for it. The Pi could monitor the soil moisture, and let me and the whole office know Bob needed attention.
Phase 1 - Messaging
The system needs to let me know when Bob needs something, I thought twitter might be a neat way to start this. I used the tweepy python library to interface with the twitter API. Tweepy is a really nice tool to set up and use. Once you have a twitter account set up, you use it to set up a twitter app, and generate an access token for tweepy to use.
Phase 2 - Sensor interfacing
We now need to actually collect some real world data from the plant. I used an analog capacitive soil moisture meter for this. The capacitive probes should be more robust than the resistive versions. To get the analog data I used an automation hat for the Pi from Pimoroni. The price was right, the library included looked very simple, and it had some extra buffered inputs and ouputs that I knew would come in handy. Impressive kit for the price, simple and robust. I sourced both of these from Digikey, my greatly preferred vendor - long live Digikey.
Phase 3 - Learning
I have data now, but I need to calibrate it over time to learn when Bob is thirsty, and when he is happy.