IoT-Edge-LLM: Edge AI Engine Analysis

A project that combines IoT sensor data analysis with edge-deployed large language models for intelligent engine vibration analysis.

This project serves as the code foundation for a two-part blog series:

• The Power of LLM models at the Edge - AWS IoT - Part 1
• The Power of LLM models at the Edge - AWS IoT - Part 2

Overview

IoT-Edge-LLM demonstrates how to deploy large language models to the edge.

Architecture

The project consists of two main components:

Frontend (React)

• Modern React application with Cloudscape Design components
• Interactive parameter controls for simulation
• Visualization of vibration data
• Real-time AI analysis results and metrics

Edge-LLM (Python)

• FastAPI backend for data processing and API endpoints
• LLM integration via Ollama for local model deployment
• WebSocket communication for streaming responses
• Data processing utilities for vibration analysis

Setup Instructions

Prerequisites

• Node.js 16+ for frontend
• Python 3.9+ for backend
• Ollama installed for local LLM serving
• A preferred LLM model (default: gemma3:1b or qwen3:1.7b)

Backend Setup

1. Navigate to the edge-llm directory:

   cd edge-llm
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Start the backend server:

   python main.py
   

   The server will run on port 8081 by default.

Frontend Setup

1. Navigate to the frontend directory:

   cd frontend
   

2. Install dependencies:

   npm install
   

3. Start the development server:

   npm run dev
   

   The frontend will be available at http://localhost:5173