A Study on the Development and Application of AI Convergence Algorithm for Object Recognition and Human Dynamics Analysis in Smart Factory

Many small and medium-sized enterprises in Korea have built smart factories with government support. However, the majority of SMEs are producing products that are difficult to automate and require a lot of manpower on behalf of large companies. In particular, it is difficult for manufacturing companies to enter smart factories because it is difficult to input real-time data. According to research needs as described above, in this study, in order to improve the fact that humans collect information by relying on visual information for more than 70%, object recognition and human dynamics analysis are used to overcome the problem of blind spots in performance aggregation. In order to achieve the purpose of the study, first, the requirements were investigated by qualitative and quantitative methods, and based on this, a prototype model was designed and implemented. Second, the implemented prototype system was applied to actual small and medium-sized manufacturing companies to prove the effectiveness of the AI convergence algorithm through qualitative and quantitative evaluation. The results of this study are expected to provide basic information necessary for calculating standard costs and building digital twins of smart factories by overcoming the limitations of SMEs’ smart factory and enabling real-time data aggregation.