Many studies have also proven that growing indoor house plants, as long as being a trend, improves health, despite the difficulty to keep them alive. The fact is that when we grow plants inside our homes, they depend 100% on us and sometimes it is difficult to know what they need.
In this talk, we will explore how to improve your plants' lives by setting up a basic plant monitoring system. This might sound complicated, but it is indeed very simple and useful: thanks to using python and jupyter notebooks. Moreover, you are going to create a machine learning pipeline from going through the steps of data labeling, selecting the framework, the model, and you will learn how to deal with the challenges that one can have in these processes.
Finally, you will see how this project is an excellent way to learn how to deal with data science challenges, at the same time that you will learn plant biology and how to implement a real-world machine learning project, and at the same time help your plant to be happier.
- You can watch the talk here
(1) This talk is beginner friendly. The main idea of this talk is to show with a practical example how we can implement machine learning at home, at the same time, how we can store and work with our own data for analysis.
(2) The background knowledge should be basic python knowledge. This proposal of talk is complementary to a previous personal project called "PLANTARIS" (https://github.com/mjmolina/plantaris).
.
├── boards
│ ├── circuit_moisture.py
│ ├── circuit_watering_simple.py
│ ├── pi_monitor.py
│ └── requirements.txt
│
└── notebooks
├── circuitpython_notebook.ipynb
├── data_analysis_sensors.ipynb
├── data_labeling.ipynb
├── prediction.ipynb
├── requirements.txt
├── temp_hum_processed.csv
└── training.ipynb
This contains the scripts used to monitor the system:
| File | Platform | Description |
|---|---|---|
circuit_moisture.py |
CPX+Crickit | Needs to be called code.py. |
Get the moisture every 60s |
||
| so the RPI can read it. | ||
pi_monitor.py |
Raspberry Pi | Listen to the CPX+Crickit system |
| via Serial Port. Additionally | ||
| it takes the environmental | ||
| temperature and humidity. | ||
Generates the temp_hum.csv file. |
circuit_watering_simple.py is an example from the talk of a system,
that is in charge of watering a plant activating a relay that enables a water
pump, according to the moisture sensor values.
Jupyter Notebooks to perform different steps on this experiments:
data_labeling.ipynb, system based on Jupyter Widgets to do an interactive labeling of all the photos that the system is taking.prediction.ipynb, ML prediction step, more comments can be found inside.training.ipynb, ML training step, more comments can be found inside.circuitpython_notebook.ipynb, configuration to use the CPX directly from a Notebook.data_analysis_sensors.ipynb, analysis of all the data gathered by the system.