Rolling bearings are an important element for industrial rotating machinery. Damage occurring on rolling bearings causes serious problems for both equipment and safety. Early detection of signs of damage on rolling bearings not only minimizes the risk of serious damage to equipment but also ensures safety for operators. In the context of today’s 4.0 industrial revolution, along with the strong development of artificial intelligence (AI), neural networks have shown their advantages in detecting and diagnosing rolling bearing damage. However, the use of traditional neural networks still has certain limitations for the need for automatic diagnosis of damage. This study presents a rolling bearing damage diagnosis algorithm based on deep learning. The study will also specifically present the clear advantages of deep learning compared to traditional neural networks. The results of the algorithm are demonstrated through experiments for fixed rotational speeds. In which, the rolling bearing vibration time signal is collected from the model through the vibration sensor.