Supply chain resilience integrating a permacrisis perspective: Multi-agent simulation and topological optimization of Tesla Giga Shanghai

Read full text Summarise with VIRAT
Authors: DongHui Duan, duan dongli
Cited by: 0
References: 25

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

In the current era of “Permacrisis,” global supply chains are transitioning from facing occasional “Black Swan” events to enduring continuous stress tests from “Gray Rhino” risks (highly probable but often neglected threats). This paper integrates Complex Adaptive Systems (CAS) theory with Agent-Based Modeling (ABM) to investigate the non-linear impact of network topology on supply chain resilience. First, by introducing triangular structures into single-chain networks, we reveal a “Redundancy Paradox”: while moderate redundancy significantly reduces the system damage ratio, excessive connectivity beyond a configuration-specific optimum (six triangular additions under the parameter setting used here) accelerates risk propagation and exacerbates instability. The existence of a U-shaped optimum is robust across parameter perturbations; its precise location is configuration-dependent. Second, taking the Tesla Shanghai Gigafactory during the 2022 Shanghai lockdown as an empirical case, we calibrate the model against the observed W-shaped (double-dip) monthly production trajectory, specifically reproducing the cascading effects of the “Aptiv wire harness crisis” through the short-board constraint of the model. The simulation results demonstrate that establishing a horizontal collaboration mechanism among homogeneous suppliers (evolving from a Star to a Mesh topology) is superior to simple connectivity increases. Under identical random shock seeds and R = 100 Monte Carlo replications, this optimization strategy reduces the time-to-80%-recovery (days required for total production value to regain 80% of its pre-shock steady-state level) by approximately 74% relative to the unoptimized star network, while the steady-state production value itself improves by a more modest ∼ 9 % — i.e. the dominant gain is in recovery speed rather than in long-run output. This study provides a theoretical basis for manufacturing enterprises to build “strategic redundancy” and survive in extreme environments.

Fields of Study & Subjects

Supply Chain Resilience and Risk Management Supply Chain and Inventory Management Infrastructure Resilience and Vulnerability Analysis

Publication details

Journal / source
Sustainable Futures
Volume
12
Pages
101938–101938
Publication date
29 October 2026
Language
EN