Final Year Project

Project Title

Predictive Fault Management in Power Distribution Networks

Description

This research addresses the energy security challenges of Karachi, a metropolis of 20 million, by transitioning utility operations from a reactive to a proactive model. Utilizing three years of K-Electric historical data, live sensor readings, and localized variables like weather and traffic, the project develops a context-aware predictive system. The primary aim is to accurately identify fault locations, classify failure types, and estimate "Time to Repair" (TTR). By integrating "Just-in-Time" maintenance and risk mapping, the system reduces operational costs, minimizing the 70 PKR cost per manual complaint, and enhances customer trust through transparent communication. Validated against actual Outage Management System (OMS) data, this model leverages unutilized logs to create a robust, localized solution for Karachi’s unique socio-economic and climatic landscape.