Applying IoT solution for Plant monitoring

Project provides codes for IoT devices, MQTT protocol, WebSocket server and Machine learning in order to monitor and analyse plant data.

Getting Started

Folders structure

plant-monitoring/
├── README.md               // Documentation file
├── iot-device              // Codes for IoT devices
├── server
│   ├── data-analysis       // Data, models and codes for data analysing
│   ├── mqtt-client         // Mqtt client which subscribes to topic published by the devices
│   └── websocket-server    // Websocket server which maintains real time connection with Mqtt client and web application
└── web-app                 // React web app for user interface

Prerequisites

Server

Necessary package

numpy
pandas
matplotlib
sklearn
keras
tensorflow
paho-mqtt
flask_socketio
socketIO-client
pymongo

Web application

Dependencies can be found in package.json file.

Installing

Server

In server directory, creating a virtual environment for deploying server:

virtualenv env -p python3

Activating virtual environment:

source env/bin/activate

Installing dependencies:

pip install {package}

Running mqtt subcribe node:

python mqtt-client/client-subscribe.py

Running web socket server:

python websocket-server/server.py

Web application

In web-app directory, installing dependencies for deploying web application

npm install

Starting web application

npm start

Starting script can be found in package.json file.

Deployment

When deploying in a real server, PM2 is a good choice for managing process.

Result

Webapp allows users to monitor sensors and control actuators. Real product where 1,2,3 are sensors, 4 is actuator, 5 is NodeMCU, 6 is App.