BIG DATA AND ENGINEERING ANALYTICS PIPELINES FOR SMART MANUFACTURING: ENHANCING EFFICIENCY, QUALITY, AND PREDICTIVE MAINTENANCE

Abstract

This study addresses the practical problem that many smart manufacturing firms deploy isolated analytics tools without coherent big data and engineering analytics pipelines, which limits gains in efficiency, quality, and predictive maintenance. The purpose is to empirically examine how the maturity of such pipelines influences three core performance dimensions in smart factories. A quantitative, cross sectional, case-based survey design was applied to 150 respondents from cloud enabled smart manufacturing enterprises, using Likert 5-point scales to measure analytics pipeline maturity, efficiency performance, quality performance, and predictive maintenance effectiveness, with firm size, industry segment, and automation level as controls. Descriptive analysis shows moderate to high maturity (PIPE mean 3.84, SD 0.62) and positive perceived outcomes for efficiency (mean 3.91), quality (3.77), and predictive maintenance (3.69), with strong reliability for all scales (Cronbach’s alpha 0.86 to 0.89). Correlation analysis indicates significant positive associations between pipeline maturity and efficiency (r 0.52), quality (r 0.47), and predictive maintenance (r 0.58, p less than .001). Multiple regression confirms that pipeline maturity is a significant predictor of efficiency (β 0.41, R² 0.34), quality (β 0.37, R² 0.29), and predictive maintenance (β 0.48, R² 0.41) after controls, while mediation tests show that predictive maintenance partially mediates the pipeline–efficiency relationship, increasing explained variance in efficiency to 0.43. These results imply that managers should treat engineering analytics pipelines as a strategic, plant level capability, prioritizing end to end data integration, high quality sensor and enterprise data, and predictive maintenance analytics to unlock scalable improvements in throughput, defect reduction, and downtime control.