DIGITAL TWINS FOR REAL-TIME BOTTLENECK ANALYSER USING PLANT SIMULATION

This paper introduces a tool integrated with Plant Simulation for automatic bottleneck detection in industrial production lines. Initially developed as a custom library, it simulated stochastic scenarios using real factory data and employed virtual sensors to identify constraints. The solution evolved into a Digital Twin, where bottleneck logic was embedded in PLCs, enabling real-time detection within the control system. This hybrid approach — combining simulation and physical control — offers fast diagnostics and supports better decision-making, such as precise resource allocation to resolve critical issues. The tool is currently deployed at the Recife (PE) and Águas Claras (RS) sites, leading to notable gains in agility, productivity, and digital maturity. The paper details the development methodology, showcases the results, and highlights how this system contributes to a more intelligent and autonomous production environment.