Greenhouse Monitor

RMIT Programming Internet of Things Assignment 1

by Jing Li and Cheng Qian

Introduction

The project once installed and configured, allows the Raspberry Pi with Sense Hat to automatically monitor ambient temperature and humidity, and compares them with preset range data and send out-of-range reading notification to user's Pushbullet device(s). Besides, it will also have the Raspberry Pi run bluetooth scanner and send out temperature/humidity readings to user's device(s), once it finds a paired device nearby.

This project also allows user to generate sensor data report to see if there is any out-of-range reading for each day. Last but not the least, user can also use this project to generate two pre-defined data visualization png files to learn more about the collected data.

Installation

1. Flash the Raspbian image provided by RMIT School of Science to the Raspberry Pi with Sense Hat.
2. SSH into the Raspbian and run mkdir Workspaces
3. Run cd Workspaces and then git clone this repository.

Dependencies

Run the following to install all required dependencies:

1. sudo apt-get install bluez bluez-firmware bluez-tools blueman
2. sudo apt-get install libatlas-base-dev
3. sudo apt-get install sqlite3
4. pip3 install pybluez
5. pip3 install requests
6. pip3 install matplotlib
7. pip3 install seaborn
8. pip3 install pandas

Configuration

Pushbullet API

1. Go to https://www.pushbullet.com/ and create an account. Then go to My Account -> Settings and Create Access Token under Access Tokens.

2. Copy the newly created token from above step.

3. SSH into Raspbian and cd Workspaces/GreenhouseMonitor

4. Run vi token.json (one can use nano to create file instead of vi, same for all similar commands in this document) and add the following to the file and make sure to replace [COPIED ACCESS TOKEN] with your access token:

   {
     "pushbullet": "[COPIED ACCESS TOKEN]"
   }

5. Download Pushbullet App and/or web browser extensions and log in with the same account.

Create SQLite Database

1. SSH into Raspbian and cd Workspaces/GreenhouseMonitor
2. Run sqlite3 greenhouse_monitor.db
3. Type .databases to verify and .exit to quit SQLite CLI.

Cron Job

1. SSH into Raspbian and cd ~
2. Run mkdir CronJobs

Greenhouse Monitor

1. vi CronJobs/RunGreenhouseMonitor.sh and add the following to the file and save:

   #!/bin/sh
   cd /home/pi/Workspaces/GreenhouseMonitor; python3 monitor_and_notify.py

2. Run chmod +x /home/pi/CronJobs/RunGreenhouseMonitor.sh

3. sudo vi /etc/cron.d/GreenhouseMonitorJob and add the following to the file and save:

   * * * * * pi /home/pi/CronJobs/RunGreenhouseMonitor.sh

Greenhouse Bluetooth

1. vi CronJobs/RunGreenhouseBluetooth.sh and add the following to the file and save:

   #!/bin/sh
   cd /home/pi/Workspaces/GreenhouseMonitor; python3 greenhouse_bluetooth.py

2. Run chmod +x /home/pi/CronJobs/RunGreenhouseBluetooth.sh

3. sudo vi /etc/cron.d/GreenhouseMonitorJob and add the following to the file and save:

   * * * * * pi /home/pi/CronJobs/RunGreenhouseBluetooth.sh
   * * * * * pi sleep 30; /home/pi/CronJobs/RunGreenhouseBluetooth.sh
   

   The script itself takes about 15 seconds to run, and the two cron jobs above run the script every 30 seconds and therefore create a 15-second interval between two runs.

Finally, run sudo reboot to restart Raspbian for the changes to take effect.

Usage

Greenhouse Monitor

Once configured, the script should automatically run every minute and it will collect the sensor data (temperature and humidity) and save them to the local database. For every reading, it will also compare with the config.json file and notify user if the reading is out of range. The notification is sent maximum once each day, local time.

To manually run the script, run:

cd /home/pi/Workspaces/GreenhouseMonitor; python3 monitor_and_notify.py

Greenhouse Bluetooth

Once configured, the script should automatically run every 30 seconds and it will look for paired devices. If a paired device is nearby, it will collect the sensor data and send notification with the data to the user. The notification is sent maximum once every minute.

To manually run the script, run:

cd /home/pi/Workspaces/GreenhouseMonitor; python3 greenhouse_bluetooth.py

Report Generation

To see if readings are out-of-range for each day, user can manually run: cd /home/pi/Workspaces/GreenhouseMonitor; python3 create_report.py

User can also modify the range in config.json before creating the report.

Data Visualization

User can manually run:

cd /home/pi/Workspaces/GreenhouseMonitor; python3 analytics.py

to generate two predefined graphs — line_chart.png and joint_plot.png.

The former shows the trends of temperature and humid, and the latter depicts the relationship between temperature and humidity along with their distributions.