This publication presents a set of computer vision methods combined with artificial intelligence tools that can be effectively applied to automate the monitoring of production workers’ behavior. The authors demonstrate how existing image analysis and object classification libraries can be leveraged to automatically generate movement trajectories of workers and other objects within a manufacturing environment. The application of the proposed methods results in the creation of heatmaps indicating worker presence in specific zones of the workspace, which may serve as a basis for production process optimization. The proposed approach is purely observational in nature, allowing employees to perform their tasks in a natural and unobstructed manner. As a result, it becomes possible to acquire reliable data on the actual course of the process, including information about the most time-consuming activities and transitions between individual stations within the production line.