Proposal for Enhancing Manufacturing Efficiency with Generative AI and Surveillance Cameras

This paper introduces a system aimed at maximizing the productivity of small and medium-sized manufacturing enterprises by improving operational efficiency through advanced image analysis. The system utilizes surveillance cameras to meticulously examine the spatial arrangement and activities of equipment and workers. Object detection and random forest techniques are employed to perform action recognition and distinguish tasks based on these actions. The extracted images are analyzed with GPT-4o to provide suggestions for task improvement. This system is particularly tested on inspection tasks and effectively guides improvements through detailed image analysis.