EMPW - Data Science

Data Processing Steps

First Step

• The admin creates a new sensor.
• The user links the created sensor to his account.
• The sesnor sends consumption data.

{
  "sensor_id": "5f1efd8e2514f800017a1fd4",
  "water_level": 126
}

Second Step

A script is scheduled to run every day and gets the consumption data for every registered sensor from the API.

[
  {
    "sensor_id": "5f1efd8e2514f800017a1fd4",
    "date": "2020-07-28T00:00:00.000Z",
    "data": [
      {
        "water_level": 126.0,
        "created_at": "2020-07-28T04:15:19.554Z"
      },
      {
        "water_level": 128.0,
        "created_at": "2020-07-28T04:30:22.698Z"
      },
      {
        "water_level": 230.0,
        "created_at": "2020-07-28T05:38:03.936Z"
      }
    ]
  },
  {
    "sensor_id": "5f1efd542514f800017a1fcc",
    "date": "2020-07-28T00:00:00.000Z",
    "data": [
      {
        "water_level": 100.0,
        "created_at": "2020-07-28T05:15:00.391Z"
      },
      {
        "water_level": 123.0,
        "created_at": "2020-07-28T05:30:02.945Z"
      },
      {
        "water_level": 13.0,
        "created_at": "2020-07-28T06:26:43.864Z"
      },
      {
        "water_level": 20.0,
        "created_at": "2020-07-28T06:50:53.828Z"
      }
    ]
  }
]

The script processes the data and returns back to the API the consumption statistics.

[
  {
    "sensor_id": "5f1efd8e2514f800017a1fd4",
    "date": "2020-07-28",
    "consumption": 104
  },
  {
    "sensor_id": "5f1efd542514f800017a1fcc",
    "date": "2020-07-28",
    "consumption": 30
  }
]

This data is stored and provided to the users.

Third step

In the frontend(web/mobile), the users sees charts and statistics for his/her water consumption.

Running the script

The script needs python3 installed.

pip3 install -r requirements.txt

cp .example.env .env # copies .example.env to .env

Open .env file and provide values for the required environment variables.

API_BASE_URL=https://api.example.com
[email protected]
ADMIN_PASSWORD=password

Run the script

python3 main.py

Deployment

The script can be deployed to Heroku to make use of Heroku Scheduler. The script runs everyday at midnight, exactly at 12:30 AM. The script requests the consumption data from the API. The API returns the consumption data of the previous day. The script processes the data and returns the consumption reports to the API.