Implementasi Diagram Kendali Shewhart untuk Pemeliharaan Prediktif Berbasis PLC dan IoT

Abstract

Predictive maintenance is a condition-based maintenance strategy that utilizes machine data to detect potential failures at an early stage before major breakdowns occur. This system integrates automation technology (PLC), the Industrial Internet, and statistical data analysis. This study discusses the application of the Shewhart control chart for automatic monitoring of servo motor torque and pneumatic cylinder cycle time. Torque and cycle time data are read by the PLC and transmitted to a server via Node-RED. On the server, the mean and standard deviation of these parameters are calculated, and the Upper Control Limit (UCL) and Lower Control Limit (LCL) are determined using a 3σ threshold. Data points falling outside the UCL/LCL are identified as process anomalies. In a case study involving several machines equipped with multiple servo motors and pneumatic cylinders, the system successfully detected servo motor torque anomalies before significant damage occurred. The analysis results indicate that with this methodology, the production process can run smoothly, since machine failures can be detected earlier, allowing preventive actions to be properly scheduled outside production hours.