Power System Fault Detection

ML project for AICTE Edunet IBM SkillsBuild Internship (July 2026).

Problem

Power distribution faults (line breakage, transformer failure, overheating) cause costly downtime. This project predicts fault type from operational parameters using machine learning.

Solution

• Model: Snap Logistic Regression trained via IBM WatsonX.ai AutoAI
• Frontend: Streamlit web app
• API: IBM Watson Machine Learning REST endpoint

How It Works

1. Enter system parameters (voltage, current, power load, temperature, etc.)
2. App calls IBM Watson ML deployed model via REST API
3. Returns predicted fault type + confidence probabilities

Setup

pip install -r requirements.txt

Create .env file:

WATSONX_API_KEY=your_api_key
DEPLOYMENT_ID=your_deployment_id

Run:

streamlit run app.py

Dataset

• 506 records, 13 columns, 3 fault types
• Source: Kaggle (CC0 license)
• Features: Fault ID, Fault Location, Voltage, Current, Power Load, Temperature, Wind Speed, Weather, Maintenance Status, Component Health, Duration, Downtime

Project Structure

power_system_faults_project/
├── app.py                 # Streamlit frontend
├── fault_data.csv         # Dataset
├── requirements.txt       # Python dependencies
├── .env.example           # Environment variable template
├── .gitignore
└── README.md

Technology

• IBM WatsonX.ai Studio AutoAI (model training)
• IBM Watson Machine Learning (deployment & API)
• Python / Streamlit / Plotly