Predictive Maintenance using IBM AutoAI

This project predicts industrial machine failures using sensor data and IBM Watson AutoAI.

Problem

Industrial machinery can fail unexpectedly, causing downtime and high costs.
Goal: Predict failures like tool wear, heat issues, and power failures using machine learning.

Solution

• Used IBM AutoAI to build and deploy a predictive model
• Trained on a Kaggle dataset of sensor readings
• Selected Snap Random Forest Classifier as the best model

Files

• Project_PPT_Predictive-Maintenance-ML-model.pdf – Project presentation
• predictive_maintenance.csv – Dataset
• README.md – Project summary

Tools & Tech

• IBM Watson AutoAI
• IBM Cloud Lite

Future Work

• Real-time IoT data integration
• Edge deployment
• Anomaly detection with unsupervised learning