Digital Twin-Enabled Supply Chain Visibility and Disruption Response in Multi-Tier Manufacturing Networks: Architecture, Implementation, and Performance Analysis

Abstract

Supply chain disruptions have emerged as a critical vulnerability for global manufacturing networks, with the COVID- 19 pandemic, geopolitical tensions, and extreme weather events exposing the fragility of lean globally integrated supply chains. This paper presents the design, implementation, and 16- month operational performance evaluation of a Digital Twin Supply Chain Architecture (DTSCA) deployed across a 78- node multi- tier manufacturing network in the electronics sector. The DTSCA integrates real- time IoT data ingestion (347 data streams), a live digital twin network model, predictive analytics based on LSTM and XGBoost algorithms, and a prescriptive MILP optimisation engine for disruption response planning. Over the evaluation period, the system detected 34 disruption precursors with an average lead time of 8.4 days before material impact, enabling proactive mitigation in 85.3% of cases. Unmitigated disruption events decreased by 68.4% relative to the pre- implementation baseline, emergency procurement spend declined by 41.2%, and on- time delivery performance improved from 71.3% to 91.7%.