Predictive Maintenance Model for AES/Eletropaulo

This project focused on developing a predictive maintenance model for AES/Eletropaulo to reduce emergency repairs and optimize maintenance processes. The objective was to predict equipment failures and prioritize preventative maintenance based on data-driven insights.

I gathered extensive data on equipment maintenance history, environmental factors, and previous failure rates. Using this data, I developed a classification model that accurately predicted potential equipment failures. The model's probability rankings allowed the maintenance team to prioritize interventions based on failure likelihood and impact, reducing the rate of maintenance-related outages from 25% to below 18%. This significantly enhanced maintenance efficiency and reduced emergency repairs.