The goal of this project is to create a smart mini greenhouse, capable of monitoring and controlling its own climate. The greenhouse will employ a user interface which will allow for remote interaction with the system. Through the interface, the user will be able to visualize the current and historical performance of the system, as well as set healthy environment variables for their plant, which the system will optimize towards. The system will utilize a custom reinforcement learning algorithm to determine which actions it should take upon each update in order to optimize its climate. The system will also include a software controller which schedules updates, carries out I/O operations, and manages sub-system processes.

This repository contains the software elements of our system, as well as Jupyter Notebook files related to the overall system.

This project is created for the NYSG Internship program at Lockheed Martin, and will serve the Engineering Explorers Program in the future.

• Python 3.6+

• Django (pip3 install Django)
• Scipy (pip3 install scipy)
• screen (sudo apt install screen)

This software is intended to be run on a Raspberry Pi running the Raspberry Pi OS (with recommended software). Development is intended to be conducted via SSH. All instructions hereafter assume that these conditions are met.

• Run 'pip3 install Django'
• Run 'pip3 install scipy'
• Run 'sudo apt install screen'

• Run 'cd ~'
• Run 'git clone https://github.com/colestrine/NYSG.git'

• Run 'cd NYSG'

• Run 'screen -S greenhouse'
• This will create a new screen session called 'greenhouse'
• Before ending our SSH session, we can detach from this screen session, and it will continue to run as a separate shell process on the Raspberry Pi

• Run './start-system.sh'
• Wait for system to start up
• Once the success message appears, you can navigate to the specified IP address in your browser to access the UI
• If your device is the only Raspberry Pi on your LAN network, you may also access the UI by navigating to 'raspberrypi:8080'

• Type 'ctrl+a', then 'd' to detach your shell session from the current SSH shell session
• This will return you to the original SSH session, and move the screen session to the background
• When we close our SSH connection, the screen process will continue to run on the Raspberry Pi

• You are now free to close your SSH connection

• Open a new SSH connection
• Run 'screen -r greenhouse'
• You will now be returned to your screen session
• Be sure to detach your screen session (see above) before closing your SSH connection to keep the system running