Abstract: In recent years, deep neural networks have shown promising results in modern fault diagnosis. This paper focuses on diagnosing actuator faults in quadcopters using a deep learning strategy.