The integration of machine learning (ML) in virtual workstations has emerged as a promising solution for improving efficiency and productivity in Industry 4.0. One of the critical environmental variables in many manufacturing contexts is humidity, which is considered a test factor for assessing the reliability and performance of processes. This work addresses the need for an ML-based ecosystem in digital manufacturing by proposing a connected system that incorporates all elements of predictive maintenance functionality in a virtual workstation. The system allows real-time monitoring and analysis of critical manufacturing processes, enabling early detection of potential issues and facilitating proactive maintenance interventions. By establishing virtual visibility of the workstation, decision-makers can remotely access and evaluate the manufacturing processes, leveraging ML algorithms to identify patterns and anomalies. This enables timely decision-making and reduces downtime, leading to increased operational efficiency and productivity. To validate the effectiveness of the generalized proposed approach, a practical example of its implementation in a relevant manufacturing facility has been considered. This integration places an emphasis on humidity as a key factor in the manufacturing process and its potential impact on product and process reliability. The generalized nature of this ecosystem justifies its applicability to various manufacturing industries, emphasizing its potential as a scalable solution for optimizing Industry 4.0 operations.