The efficient transmission of electrical power is essential for the stability and reliability of modern grids. Insulators play a critical role in isolating high-voltage conductors, yet they are highly susceptible to defects such as cracks, contamination, and mechanical damage, which can cause power losses, equipment failure, and outages. This study presents an Unmanned Aerial Vehicle (UAV) based inspection framework integrated with the You Only Look Once version 8 (YOLOv8) deep learning model for automated detection of insulator faults. A custom UAV platform was developed to ensure stable flight, adaptability, and reliable data acquisition under diverse operational and environmental conditions. Using this platform, a dedicated dataset of 8,300 high-resolution annotated images of transmission insulators was collected across varying weather and lighting scenarios. To improve generalization and mitigate overfit-ting, advanced data augmentation strategies were employed. The proposed system achieved a detection accuracy of 92% and a mean Average Precision (mAP@[0.5:0.95]) of 50.1%, outper-forming benchmark methods such as R-FCN, Faster R-CNN, YOLOv5, and YOLOv7. These results confirm the novelty and effectiveness of combining UAV-based aerial imaging with advanced deep learning for reliable, real-time insulator monitoring. The approach reduces dependence on manual inspections and enables predictive maintenance, thereby contributing to improved grid reliability, operational efficiency, and cost-effective infrastructure management.