IOT Dashboard

Problem to solve

With the COVID-19 pandemic, we were all forced to work from home. This brought a whole new set of new hobbies and problems with it, that we try to address with this IoT project.

Devices

Philips Hue

Philips Hue is a system that consists of LED lamps, sensors, and switches. The system can be controlled manually using a Smartphone or can be automated, for example using Apple HomeKit.

Xiaomi Miflora

The Miflora is a wireless bluetooth sensor platform that allows to monitor the condition of plants. It monitors the moisture, ambient light, and temperature.

Xiamoi Temperature and Humidity Monitor 2

The Temperature and Humidity Monitor 2 is a small bluetooth device with an integrated display. It can display the room temperature and humidity.

Files

In the following sections all files and their use are described.

configuration.yaml

An example configuration file example.configuration.yaml is provided.

configuration.py

Abstraction layer for the configuration.yaml file that contains all the config items for this project. This file handles loading and accessing the configuration file.

flower_status.py

Reads the data from one Xiaomi Miflora plant sensor and publishes the temperature, moisture, and light data as a JSON object to a message queue in the topic sensor/flower. JSON format:

{
  "temperature": 12,
  "moisture": 35.7,
  "light": 150
}

room_monitor.py

Reads the data from one Xiaomi Temperature and Humidity monitor 2 and publishes the temperature and humidity data as a JSON object to a message queue in the topic sensor/room. JSON format:

{
  "temperature": 26.2,
  "humidity": 42,
}

hue_light_status.py

Reads the data from Philips Hue light system and publishes the light groups with their name and state (on/off) data as a JSON object to a message queue in the topic sensor/hue/light. JSON format:

[
  {
    "name": "Living Room",
    "state": true
  },
  {
    "name": "Hallway",
    "state": false
  }
]

hue_light_control.py

This script can turn Philips Hue lights on and off depending on their current light status. The script expects a json object as string as the first parameter where the group name is saved in the name property of the json object. The group name has to be converted to the interal Hue id to control the lights of the group.

hue_motion_sensor.py

Reads the motion sensor data from the Philipps Hue bridge and publishes the sensor information to MQTT with the information presence_detected to the topic sensor/hue/motion. The event will only be published when presence is detected. Otherwhise nothing will be sent.

{
    "presence_detected": true
}

mitemp_monitor2_poller.py

Class that can read sensor data from a Xiaomi Temperature and Humidity Monitor 2 based on a mac address. Returns an object with the attributes temperature and humidity.

node_red_flow.json

This file contains the node-red flows that are used to build the dashboard and read the data.

rain_check.py

Reads the weather data from pyowm and checks the next hours for rain. The output will be written to stdout for further processing in node-red. JSON format:

{
  "will_rain": true
}

umbrella_detection.py

Fetches one image from the attached webcam and executes an object recognition search using tensorflow. A SSD model trained on the COCO dataset is used for the object recognition. The result will be written to stdout for further processing in node-red. JSON format:

{
  "umbrella_detected": true
}

send_umbrella_reminder.py

Python script that interacts with Twilio to send a Whatsapp message that reminds the user that he/she has forgotten the umbrella.

weather_forecast.py

Reads the weather data from pyowm and gets the forecast for the next 5 days. The output will be written to stdout for further processing in node-red. JSON format:

[
  {
    "date_unix": "12345",
    "date_formatted": "17.12",
    "min_temperature": 0,
    "max_temperature": 10,
    "status": "cloudy",
    "detailed_status": "cloudy overcast"
  },
  {
    "date_unix": "12345",
    "date_formatted": "18.12",
    "min_temperature": 0,
    "max_temperature": 12,
    "status": "cloudy",
    "detailed_status": "cloudy overcast"
  }
]

Used Libraries

The following libraries were used for this project.

Library	Version	Use case
bluepy	1.3.0	Python library to interface with the bluetooth system on Linux.
btlewrap	0.0.10	Bluetooth wrapper library for different bluetooth backends.
pyowm	3.2.0	Library to read weather information from OpenWeatherMap (OWM). Used to check for rain and get the weather forecast.
phue	1.1	Used to read the data from the Philipps Hue light system.
miflora	0.7.1	Used to read sensor data from the Xiaomi Miflora plant sensor.
tflite_runtime	2.5.0.post1	AI library to run object recognition on a webcam image with the aim to detect if an umbrella is present or not. Runtime/Light version of tensorflow
numpy	1.16.2	Used to transform the image for tensorflow.
opencv	4.1.0.25	Fetch the picture from the attached webcam.
paho_mqtt	1.6.1	MQTT library for Python
requests	2.21.0	Library to execute HTTP requests
twilio	7.4.0	Library to send whatsapp or sms