Skip to content

Latest commit

 

History

History
104 lines (77 loc) · 6.69 KB

installation.md

File metadata and controls

104 lines (77 loc) · 6.69 KB
Raw

SYSC3010 Project Installation Instructions

The Floralyfe project comprises a distributed remote plant monitoring solution that is composed of three core components: a frontend React client, a backend Node server and distributed Raspberry Pi systems.

Development

At a high level, the Floralyfe system consists of three core nodes: the client, the server and the pi system. The server serves as the communication layer, routing requests between the physical plant systems and the client. It also abstracts the cloud database. The client serves as the user's interface to the system while the pi system serves as the actual plant monitor and irrigation solution. To get these nodes running, follow the instructions below.

pi

The Raspberry Pi node is written in Python. It is recommended that you setup a virtual environment to run the python instance along with its dependencies. Follow the instructions below to set up a local environment.

Proceed to the /pi directory and execute the following commands depending on your operating system.

  1. Ensure that the python virtualenv module is installed on your machine by using pip install virtualenv.

Windows

  1. Initialize a virtual environment called env using python3 -m venv env. Alternatively, .\venv.bat will do this for you.
  2. Once the environment has been created, you should see a env directory created under the pi/ root. To enter the virtual environment, activate it using .\env\Scripts\activate.
  3. Within the environment, run pip install -r requirements.txt to install dependencies. Alternatively, .\scripts\install.bat will do the same thing.
  4. To exit the virtual environment, run deactivate.

Note: Scripts for mounting/demouting the virtual environment could not be created without provided absolute paths to the activate and deactivate scripts.

Linux (Raspbian)

  1. Scripts have already been provided for you to simplify the process. Feel free to open them to see the commands being executed.
  2. Initialize the virtual environment by running source scripts/venv.sh.
  3. Once initialized, an env/ directory should have been created under pi/.
  4. Ensure you've entered the environment (there will be a (env) prefix before your working directory). If you're not in the director, see step 5. Install all dependencies using source scripts/install.sh.
  5. If the virtual environment is already installed, run source scripts/run.sh to enter it. The command source scripts/exit.sh is used to exit the environment.

Note1: By default, you will not be able to interface with the SenseHat. Follow these instructions to resolve this issue

Note2: To stabilize the servo pwm, we use a module named pigpio. Follow these instructions to install it: https://abyz.me.uk/rpi/pigpio/download.html. Other pi specific modules that must be installed: busio, digitalio, board and adafruit_mcp3xxx.mcp3008 (some may already be installed on your pi).

Note3: The location of all script files is under pi/ and they must be referenced accordingly. This means you should always call scripts/script.* rather than moving within the scripts/ directory and calling a script directly.

Once the environment and dependencies have been setup. Navigate to src/ and execute the entrypoint python main.py to run the pi instance. Refer to the notes below for additional development details.

Additional Notes

  • The pi node implements a unit test workflow, to help develop code that passes these tests, follow these instructions to setup linting for VSCode.

  • Tests can be executed using the srcipts/test.sh or .\scripts\test.bat scripts. Alternatively, use the pytest -v --mocha test/ command in src. *All unit tests should be added to the test/ directory with a test_ prefix for both the file and test functions within. Tests can be ignored with the --ignore=<file_path> argument.

  • Attention: Standalone testing of modules must be performed from within src/ such that src/ remains the root of the application and imports keep working. See camera_standalone.py and irrigation_standalone.py for examples.

  • Run mypy --strict --implicit-reexport $(git ls-files '*.py') before making any pull requests and ensure all tests pass.

  • Anytime you install a module, run pip freeze > requirements.txt to save it.

  • Note: Make sure to change the following package versions to avoid installation errors in the future:

numpy==1.19.5
opencv-python==4.5.3.56
tomli>=1.0
websockets==9.1
Pillow==8.4.0
scipy==1.5.4
pkg_resources==0.0.0        # remove this
picamera==1.13              # remove this
RTIMULib==7.2.1             # remove this

server

The server node is written in TypeScript and manages communication accross application nodes. The node exposes a GraphQL API to encapsulate the database layer and also exposes an explicit WebSocket Server. The dedicated WebSocket has been setup for client <-> system messaging in addition to pubsub based WebSockets abstracted by GraphQL subscriptions.

There are several services used by the server that require corresponding API keys. You'll need to create a .env file in the server root that looks like this:

EXPRESS_PORT=5000
GRAPHQL_PORT=5001
DEVELOPMENT=1

# Firebase Authentication Credentials
FIREBASE_API_KEY=
FIREBASE_AUTH_DOMAIN=
FIREBASE_PROJECT_ID=
FIREBASE_STORAGE_BUCKET=
FIREBASE_MESSAGING_SENDER_ID=
FIREBASE_APP_ID=

# Notification Email Sender Credentials
EMAIL_USER = # The email address
EMAIL_PASS = # The email password

Once that's done, you can initialize and start the server with the following commands:

  1. Run npm i to install all dependencies
  2. Run npm run emulator-test to run unit tests (that require a Firestore emulator).
  3. Run npm run dev to start the server in development on port 5000.

client

The client is based on NextJS+ChakraUI and is used as the user's primary interface to the system. It is subscribed to messages sent to the logged in user at the WebSocket and GraphQL subscription channels and can also send messages through the WS connection.

The client uses Plant Id to provide plant recognition. To enable this feature, you'll need to create a .env.local file in the client root with the following.

# PlantID
NEXT_PUBLIC_PLANT_ID_API_KEY=
NEXT_PUBLIC_DEVELOPMENT=true
  1. Run npm i to install all dependencies
  2. Run npm run dev to start the client in development on port 3000.

Additional Documentation

OpenCV Luminescense Calculation