Paper Title
GIS Driven Transformer Health Monitoring System

Abstract
Power transformers are critical assets in modern power distribution networks, and unexpected failures can lead to outages, equipment damage, and costly maintenance. Traditional monitoring techniques rely mainly on manual inspection and offline diagnostic tests, which offer limited real-time visibility into transformer health. This paper proposes a GIS-Driven Transformer Health Monitoring System that integrates IoT-based multi-parameter sensing, machine-learning-based health classification, and geospatial visualization for intelligent asset management. IoT sensors measure temperature, oil level, vibration, humidity, current, and voltage, and the collected data is processed using a cloud-based Random Forest classifier that categorizes transformer health into low, medium, and critical risk levels. A GIS dashboard provides real-time spatial visualization of transformer health across locations, enabling utilities to identify high-risk regions and optimize maintenance activities. While the current evaluation uses a controlled dataset, the system architecture demonstrates significant potential for supporting predictive maintenance and improving smart grid reliability. Future work will incorporate field data and advanced deep-learning models for enhanced robustness. Keywords - IoT, GIS, Machine Learning, Transformer Monitoring, Random Forest, Smart Grid