Enhancing Safety Notification and Invoice Management Processes through Integration of Robotic Process Automation (RPA) and Artificial Intelligence A Case Study of Siam Denso Manufacturing Co., Ltd.

Abstract

This study aimed to: 1) examine and analyze the existing manual workflow for safety-incident notification (Yokoten) and invoice management in order to identify bottlenecks and human errors; 2) develop and integrate an automated system combining Robotic Process Automation (RPA) with Artificial Intelligence (AI) for data extraction, language translation, and systematic data entry into the Enterprise Resource Planning (ERP) system; 3) evaluate  the efficiency of the automated invoice-upload system in terms of operational time reduction and data accuracy, and gather user feedback; and 4) propose guidelines for digital transformation in the automotive industry. Four key informants directly involved in the safety-management and accounting processes at Siam Denso Manufacturing Co., Ltd. were purposively selected to evaluate the system. The system was developed primarily with UiPath for workflow orchestration, together with Microsoft Copilot Studio and Optical Character Recognition (OCR) for translation and data extraction; data were collected through pre/post time measurements and an informal user-opinion survey and analyzed using percentages and qualitative description. The results showed that the automated invoice-upload system reduced redundant manual workload in line with the target of 200 person-hours per year,    and the user-opinion survey indicated overall satisfaction at 100 percent of respondents. These findings suggested that integrating RPA with AI could reduce routine documentation workload and human error, offering an initial pathway toward digital transformation for organizations in the automotive industry.