Digital-Twin And IOT-Based Condition Monitoring for Rotating Machinery and Process Equipment in Petrochemical and Industrial Environments

Abstract

Digital Twin and Internet of Things (IoT)-based condition monitoring have emerged as essential technologies for improving the reliability, maintenance effectiveness, and operational performance of rotating machinery and process equipment within petrochemical and industrial environments. This study quantitatively examined the influence of Digital Twin capability, IoT monitoring infrastructure, equipment condition monitoring effectiveness, and predictive maintenance performance on operational performance using an integrated conceptual framework grounded in Digital Twin Theory, Cyber-Physical Systems Theory, Condition-Based Maintenance Theory, Reliability-Centered Maintenance Theory, Predictive Analytics Theory, and Systems Theory. A quantitative cross-sectional correlational research design was employed. Data were collected from industrial professionals working in petrochemical, oil and gas, chemical manufacturing, and heavy industrial organizations using a structured five-point Likert-scale questionnaire. A total of 420 questionnaires were distributed, 389 responses were received, and 368 valid questionnaires were retained for statistical analysis, yielding an effective response rate of 87.62%. Data analysis was conducted using SPSS and AMOS through descriptive statistics, reliability analysis, Confirmatory Factor Analysis, Pearson correlation, multiple regression analysis, and Structural Equation Modeling. The measurement model demonstrated excellent psychometric properties, with Cronbach's alpha values ranging from 0.918 to 0.941, Composite Reliability values ranging from 0.931 to 0.949, and Average Variance Extracted values between 0.702 and 0.758. The structural model also demonstrated satisfactory goodness-of-fit, including χ²/df = 2.184, CFI = 0.972, TLI = 0.968, GFI = 0.948, RMSEA = 0.057, and SRMR = 0.041. Multiple regression analysis indicated that equipment condition monitoring effectiveness was the strongest predictor of operational performance (β = 0.374, p < 0.001), followed by predictive maintenance performance (β = 0.289, p < 0.001), Digital Twin capability (β = 0.248, p < 0.001), and IoT monitoring capability (β = 0.192, p < 0.001). The integrated model explained 79.6% of the variance in operational performance (R² = 0.796), while the predictive relevance value reached 0.581 and overall predictive accuracy was 91.8%.