In the pursuit of enhanced productivity, reduced costs, and minimized lead times, manufacturers are transitioning from traditional systems to autonomous systems. This shift, driven by the emergence of smart manufacturing and technological advancements such as robotics, collaborative robots (Cobots), automation, and digitalization, necessitates a parallel evolution in safety protocols—termed Safety 4.0—to mitigate the risks associated with such dynamic environments. The integration of smart technologies within manufacturing significantly transforms traditional workflows and intensifies the need for comprehensive safety training and guidelines. Innovations like smart personal protective equipment (PPE) and wearable sensors are pivotal in this transition, yet they often prove financially burdensome for manufacturers due to high costs and the scale of workforce deployment. Moreover, the effective use of these technologies requires continuous monitoring and data analysis, further straining resources. To address these challenges, this paper proposes the adoption of computer vision technology to enhance safety measures within manufacturing facilities, focusing on human and PPE detection. It details a holistic methodology encompassing data collection, preprocessing, training, and execution. The discussion extends to the implementation framework of this technology, emphasizing its role in enabling autonomous decision-making—a crucial step beyond mere detection. Furthermore, the paper explores the utilization of the accumulated data to develop immersive training modules employing Mixed Reality, thereby reinforcing safety protocols and fostering an environment of continuous learning and adaptation. This approach not only contributes to safeguarding personnel but also aligns with the financial and reputational interests of forward-thinking manufacturers.