The main objective was to demonstrate how automated video analysis can effectively identify muda in industrial settings processes within the principles of the Lean Management framework.

To achieve this, the thesis defined and tackled research questions focusing on the type of muda detectable via the proposed solution, the integration of machine learning and large language models to improve the detection accuracy and scalability, and the challenge and solution for the existing video analysis methods in a Lean context.

Key findings highlights that the proposed solution of combining an automated system with Lean Management principles can reduce reliance on subjective manual analysis. Therefore, improving the scalability and precision of muda detection. This dissertation establishes a framework for future practical implementation of automated video analysis in industrial settings that aligns with Lean principles to improve operational efficiency and foster an organizational culture of continuous improvement.