Design and Implementation of an AI-Driven Industrial Production and Supply-Chain Optimization System

Abstract

The increasing integration of artificial intelligence into industrial production and supply-chain operations has created opportunities for more intelligent planning, forecasting, maintenance, sourcing, logistics, and decision-making, yet many industrial organizations continue to experience fragmented information systems, uncoordinated AI applications, unexpected equipment failures, forecasting inaccuracies, inventory imbalances, supplier uncertainty, and operational disruptions. This study aimed to design and empirically evaluate an integrated AI-driven industrial production and supply-chain optimization system by determining how multiple AI-enabled capabilities individually and collectively contribute to Industrial Production and Supply-Chain Optimization (IPSCO). A quantitative, cross-sectional, case-study-based research design was employed using enterprise-level industrial and manufacturing cases represented by professionals working across production, operations, engineering, maintenance, procurement, inventory, logistics, information technology, data analytics, and AI-related functions. Of 300 questionnaires distributed, 281 were returned and 270 valid responses were retained, producing a usable response rate of 90.0%. The principal variables were AI-Driven Production Planning and Scheduling (AIPPS), AI-Enabled Predictive Maintenance and Process Intelligence (AIPMPI), AI-Driven Demand Forecasting and Inventory Optimization (AIDFIO), AI-Enabled Supplier, Procurement, and Logistics Intelligence (AISPLI), Real-Time Industrial Data Integration and AI-Supported Decision-Making (RTIDAIDM), and IPSCO. Analysis incorporated descriptive statistics, Cronbach’s alpha reliability testing, Pearson correlation, multiple regression, capability profiling, and production-supply chain synchronization analysis. IPSCO recorded M = 4.18, SD = 0.54, while RTIDAIDM achieved the highest AI capability mean, M = 4.15, SD = 0.55. All predictors were significantly associated with IPSCO, with RTIDAIDM showing the strongest correlation, r = .72, p < .001. The regression model explained 69.9% of IPSCO variance, R = .836, R² = .699, adjusted R² = .693, F (5, 264) = 122.59, p < .001. RTIDAIDM was the strongest predictor, β = .30, followed by AIPPS, β = .24, AIDFIO, β = .21, AIPMPI, β = .18, and AISPLI, β = .15. All six hypotheses were supported, while a synchronization gap of only 0.05 indicated highly synchronized production-side and supply-chain-side AI capabilities. The findings imply that industrial organizations can achieve stronger optimization by integrating real-time data, intelligent planning, predictive maintenance, forecasting, inventory, procurement, supplier, and logistics capabilities within a coordinated AI architecture.