Industrial internet platform-driven decentralised multi-level synchronised reconfiguration of assembly supply chains

Abstract

<p>Disruptive environments, characterised by frequent contingencies, significantly challenge supply chain operations, particularly assembly supply chains with complex structures and decentralised decision-making. Failure to promptly reconfigure chains during disruptions may ripple adverse effects through the network, causing substantial losses. Consequently, decentralised dynamic reconfiguration of assembly supply chains under contingencies warrants attention. However, existing research on this topic is limited, necessitating further exploration utilising emerging technologies. Isolated information systems, such as MES and ERP, are widely deployed across firms and increasingly integrated into industry-level Industrial Internet Platforms (IIPs) to enhance information sharing and utilisation. Nevertheless, limited research explores the integration of real-time IIP data with reconfiguration mechanisms for resilient, flexible, and decentralised chain reconfiguration. Thus, this paper proposes an IIP-driven synchronised reconfiguration of supply chains (SyncRSC) solution inspired by the concepts of synchronisation and reconfigurable supply chains. SyncRSC employs the IIP architecture for real-time information support, a three-state mechanism as the qualitative method, and improved graph neural networks alongside augmented lagrangian coordination as quantitative methods. A case study of an air-conditioning supply chain verifies the superiority of the proposed methods and analyzes SyncRSC's performance under different disruption levels.</p>