Smart-Plug Energy Monitoring Service

Freelance client · Freelancer

PythonFlaskCeleryRedisMongoDBJWTSwitchBot APIMatplotlib

What it does

Turns the SwitchBot Plug Mini cloud API into a consumption history for each appliance. Users create an account and link their SwitchBot Cloud credentials. The service then discovers their plugs, records voltage, current, power and on-time every 30 minutes, and shows each appliance's energy use with charts.

How it works

  1. Signed API access. Every request to the SwitchBot v1.1 API carries a signature: Base64(HMAC-SHA256(secret, token + millisecond timestamp + nonce)), with a fresh UUID nonce for each call.
  2. Device discovery. Linking credentials triggers discovery of every plug on the account. Each plug is registered with an empty time series.
  3. Scheduled polling. A Celery beat schedule, with Redis as the broker, wakes every 30 minutes. A worker polls every plug's status endpoint for line voltage, current, power and daily on-time, appends the readings to that device's series in MongoDB, and derives energy in watt-hours from power and on-time.
  4. Two services, one codebase. A JSON REST API secured with flask-jwt-extended handles accounts, credential linking, device refresh and data retrieval. A separate server-rendered front end uses Flask-Login, with the API's JWT as the session identity. It lists each user's plugs with their accumulated energy and draws voltage, current and energy charts on demand. An app-factory structure with Celery task-context integration lets the API, worker and scheduler share one configuration.
  5. Testing without hardware. A "dummy" credential mode generates synthetic plugs and readings. The whole pipeline (sign-up, discovery, scheduled polling, charting) can be exercised without a single device.