Initial Setup and Phase
Our Energy Management System (EMS) utilizes a three-tier IoT architecture designed to decouple high-frequency sensor data from user Management.
The System Architecture is organized into three distince zones:
-
The Edge Layer (Ingestion): A Raspberry Pi polls the Inverter/Meter via Modbus RTU and pushes telemetry to the cloud every second via a persistent WebSocket connection.
-
The Cloud Layer (Processing): We utilize a 'Polyglot Persistence' Strategy. FastAPI acts as the high-perfomance hub, routing live data directly to TimescaleDB (Time-series) for efficient storage. In parallel, Django manages user identifies in SQLite and queries TimescaleDB to generate historical performance reports.
-
The User Layer (Visualization): A Next.js frontend connects to both backends receiving live updates from FastAPI via WebSockets and fetching secure historical data from Django via REST APIs.