-
Notifications
You must be signed in to change notification settings - Fork 0
Software Architecture
GiZano edited this page Jan 22, 2026
·
1 revision
This project integrates Java and Python to create a seamless control and analysis interface.
The User Interface is built with Java 23.
- WebViewer Component: Connects to ThingsBoard (port 8080) via HTTP. It renders the dashboard exactly like a standard browser.
- Console Component: Acts as a bridge to the "Python AI Analyzer".
-
Features:
- Direct browser access button.
- Theme switcher (Light/Dark).
- Console toggle (Hide/Show).
The intelligence of the system is powered by a local Large Language Model.
- Data Retrieval: The script asks for the Device ID and fetches historical data from the ThingsBoard database.
- User Input: The user selects metrics and time lapse for analysis.
-
AI Processing:
- The script connects to the local Ollama instance.
- It sends a prompt containing the data to Mistral:7b.
- It asks for specific key aspects (efficiency trends, anomalies).
- Output: The LLM's analysis is returned and displayed on the Java Console.