Predictive-Maintenance-for-Manufacturing-Machines

Project Name: Predictive Maintenance for Manufacturing Equipment

Description: Developed an end-to-end machine learning solution to predict potential machine failures using sensor data (IoT). The system monitors real-time metrics such as temperature and vibration to forecast anomalies before they lead to breakdowns.

Key Responsibilities & Tech Stack:

Data Processing: Cleaned and processed high-frequency time-series sensor data. Performed advanced feature engineering using Rolling Statistics and Lag Features.

Modeling: Built and tuned classification models (XGBoost / Random Forest) to detect failure patterns with high precision.

Challenges Solved: Addressed the class imbalance problem using SMOTE and cost-sensitive learning to improve failure detection rates (Recall).

Business Impact: Designed the model to reduce unplanned downtime and optimize maintenance schedules, translating technical predictions into actionable business insights.

Tools: Python, Pandas, Scikit-learn, XGBoost, Matplotlib