Fault Diagnosis using eXplainable AI: A transfer learning-based approach for rotating machinery exploiting augmented synthetic data

Lucas Costa Brito, Gian Antonio Susto, Jorge Nei Brito, Marcus Antonio Viana Duarte

Abstract: Highlights•FaultD-XAI: a transfer learning approach based on augmented synthetic data.•Possibility of interpreting the fault diagnosis obtained by the model.•A new contribution to Explainable Artificial Intelligence in rotating machinery.•Industrial application with the possibility of generating varied training datasets.•A New dataset publicly available: https://data.mendeley.com/datasets/zx8pfhdtnb/1.