This paper’s significant research contribution lies in the development of an original machine learning-based approach to automate insulator detection and management for overhead transmission lines. By leveraging retrained convolutional neural networks (YOLO), the study addresses challenges such as diverse image conditions and unbalanced datasets, achieving an f1-score of 97.5%. In addition to enhance the insulator detection and classification performance, we integrate our original approach seamlessly with existing asset management systems, improving real-time decision-making and reducing reliance on manual audits. This innovation significantly streamlines the maintenance and reliability of power transmission networks.