The amount of data collected as part of production processes has increased due to the proliferation of recognition technologies and the growing integration of smart systems. When processed and analyzed, this data can provide valuable information and insights from production, helping optimize operations. In the industry, the management of physical assets is critical as it directly affects the efficiency and overall performance of the factory. Therefore, asset failures need to be detected and corrected as early as possible to avoid costly downtime in production. This paper presents a machine learning model based on Industry 4.0, which is specifically designed to prevent physical asset failures in the factory. A framework utilizing a probabilistic approach was outlined to monitor the progressive wear of equipment and assist personnel in planning maintenance activities with greater insight and precision. This Bayesian model provided a useful basis for the method, highlighting its main results, challenges, and opportunities. Additionally, it serves as an essential tool to support further research in the field of predictive maintenance, fostering advancements in industrial reliability and efficiency.