INTELLIGENT CAPACITY SUPERVISION AND CONTROL SYSTEM IN INDUSTRIAL PROCESSES: INTEGRATION OF SCADA, AI, AND MACHINE LEARNING

Abstract

This study proposes a machine learning-based system for capacity supervision and control in industrial automation. The solution integrates high-precision sensors, programmable logic controllers (PLCs) and a SCADA (Supervisory Control and Data Acquisition) system, allowing real-time monitoring and adjustment of manufacturing processes. The methodology included the development of a software in C# in the Visual Studio 2015 environment, with an interface in a Mitsubishi CPU Q03UDV PLC, and the implementation of the system in a production line for practical evaluation.

The results demonstrated the system's ability to maintain the process capability indexes (CpK) above the critical limits (1.33) through the automatic correction of deviations. Key highlights include efficient integration with industrial networks and dynamic adaptation to production variabilities. On the other hand, limitations were identified, such as the dependence on a robust infrastructure and challenges in environments with high electromagnetic interference.

The discussion highlights the potential for scalability, application in other industrial contexts, and the inclusion of advanced algorithms, such as neural networks, to enhance predictive capacity. Future work suggests exploring more affordable implementations for small and medium-sized businesses, integration with IoT for predictive maintenance, and sustainability assessments. This research contributes to the advancement of intelligent automation, promoting consistent quality and operational efficiency in manufacturing.