A FOUR-LAYER MODULAR FRAMEWORK FOR KPI MEASUREMENT AND OPTIMIZATION IN SMART MANUFACTURING: A CONCEPTUAL MODEL FOR TRANSITIONAL ECONOMIES
Tanja Trifunović, Aleksandar Đorđević, Miladin Stefanović
Abstract: Industry 4.0 has opened the possibility of moving from reactive KPI monitoring toward predictive and even prescriptive performance management in manufacturing. Most existing optimization models, however, presuppose a high degree of digital maturity, tight OT/IT integration and ready access to specialized staff. These conditions are rarely met in economies still undergoing structural transition, where many factories continue to operate with fragmented systems and limited analytical capacity. This paper develops a four-layer framework intended specifically for such settings. The layers address successive stages of capability: sensor-based data acquisition supported by 5G, hybrid Edge/Cloud processing, AI/ML models that generate both forecasts and actionable recommendations, and a closed feedback mechanism that continuously refines the system. Because the architecture is modular, firms can begin with basic real-time data collection and add further layers only when resources and skills allow. The principal contribution lies in offering a staged pathway that lets smaller manufacturers start digital transformation without committing to large-scale, high-risk investments from the outset.
Keywords: Industry 4.0, Smart manufacturing, KPI optimization, Modular framework, AI/ML analytics, Transitional economies
DOI: 10.24874/PES08.03B.016
Recieved: 10.02.2026 Revised: 20.05.2026 Accepted: 10.06.2026
UDC:
Reads: 4 
