Computer Vision-Based Work Productivity Optimization with Lean Principles and Simulation Method
摘要
With the development of construction automation technologies, semi-automated production lines have become increasingly common on the construction site. Current construction productivity is affected by worker efficiency, the utilization rate of machines, and the cooperation between workers and machines. In this situation, lean thinking integrating with computer vision may be one of the most effective methods to improve work productivity further. This study aims to acquire actual work data using computer vision methods and explore the effectiveness of lean methods by identifying potential wastes in the construction process. The lean model, subsequently, is designed with Simphony, considering vision-based results like waiting time and applying lean methods like takt time and just-in-time. Finally, the study compares the simulation results, such as duration and resource utilization rate, between the traditional and lean scenarios to clarify whether the new lean production workflow is beneficial to optimizing the construction process.