Analyzing Microservice Communication Performance in Industrial Edge Platform

The integration of legacy manufacturing systems into Industry 4.0 ecosystems necessitates architectures that balance interoperability and performance across infrastructures. This paper addresses these challenges by proposing a hybrid system leveraging Docker containers, the Moleculer framework, and the Asset Administration Shell (AAS) to enable semantic interoperability while quantifying communication latencies in bidirectional industrial workflows, analyzing microservice communication performance in edge platforms through a hybrid architecture that combines legacy protocols with edge virtualization, and contextualizing it against fieldbus-based systems and fully virtualized containerized microservices. Experimental validation on a Siemens S7-1500 Programmable Logic Controller (PLC) evaluates latency distributions across architectures, revealing average round-trip times of 7.23 ms for hybrid sensor updates and 5.94 ms for virtualized command writes, with 75% of operations under 10 ms. Performance tradeoffs emerge between protocol translation overhead in hybrid systems and transporter-induced variability in virtual deployments. By correlating temporal metrics with architectural choices, this work provides actionable insights for modernizing brownfield installations, enabling manufacturers to select optimal strategies based on legacy constraints and real-time requirements. The results highlight the viability of gradual migration paths, preserving existing investments while adopting Industry 4.0 capabilities through performance-aware microservice orchestration.