Towards digital twin-enhanced control policies: A knowledge-based classification of release and dispatching policies in manufacturing systems

Abstract

The management of modern discrete manufacturing systems is challenged by high levels of complexity arising from intricate interdependencies among processes and the need to adapt to frequent internal and external disruptions. In this context, control policies play a pivotal role in managing manufacturing operations, guiding decisions for governing systems and optimising their performance. This study investigates the design and classification of release and dispatching policies based on the type and structure of information they require, with particular emphasis on supporting real-time and adaptive decision-making. This analysis takes advantage of the concept of Digital Twin (DT), tightly integrated with the physical manufacturing system via IIoT technologies, enabling continuous monitoring of operations in a factory, but also forward-looking simulation of system behaviour. The proposed classification leverages an ontology-based data model that formalises the structure of manufacturing knowledge and facilitates the systematic identification of the information required by control policies. The classification scheme incorporates both the informational requirements and the potential role of the DT in supporting their execution and the results provide a structured perspective on how control strategies can be aligned with available data and digital infrastructure to enhance the management of manufacturing systems.