Analisis Penerapan Optical Character Recognition (OCR) pada Proses Dokumen Pengiriman Outbound di PT. XYZ Menggunakan Pendekatan Learning Curve

Abstract

PT. XYZ has undergone a system transition in the Outbound Shipiing Document (DPO) process, shifting from a manual system to a digital one with the implementation of Optical Character Recognition (OCR) technology. The company has taken this step to reduce human errors in filling out checklist forms and checking the number of DPOs. The impact of human errors is the repetition of activities, resulting in time wastage. In the implementation of OCR, six employees responsible for creating DPOs are not yet accustomed to operating the digital system. Therefore, they require adaptation to the new workflow, and there is currently no standard time set to assess worker productivity. The methodology employed in this research is quantitative descriptive analysis, and the analytical method used for learning and adaptation to the new system is the learning curve analysis. The results of the learning curve analysis indicate that performance improvement or efficiency increases with experience and frequency of performing the tasks. The impact of transitioning from a manual to a digital system with OCR is a decrease in delays compared to the previous month, which were caused by human errors in DPO creation, as well as a reduction in the time required for DPO creation. Workers require varying repetitions to achieve stability levels, adjusted to each worker's learning rate. Worker 2 is used as a reference for determining the standard time, as they have the best performance with a learning rate of 92.63% requiring a total of 85 repetitions.