Human–AI Collaborative Neuromorphic Digital Twins with Adaptive Game-Theoretic Intelligence for Fuzzy Multi-Objective Optimization of Multi-Stakeholder Supply Chains

Abstract

Multi-stakeholder supply chains require decision mechanisms capable of simultaneously interpreting dynamic system states, coordinating conflicting stakeholder interests, and managing uncertainty across interconnected operational decisions. This study proposes an integrated Human–AI Collaborative Neuromorphic Digital Twin framework in which synchronized supply-chain data are transformed into cognitive representations, enriched through human–AI collaborative intelligence, strategically coordinated through adaptive game-theoretic interactions, and subsequently mapped into a unified decision-knowledge representation for fuzzy multi-objective optimization. The principal innovation lies in this closed and interconnected decision architecture, where the outputs of cognitive, collaborative, and strategic intelligence layers are explicitly fused and transferred to the optimization space rather than being applied as independent analytical modules. The framework jointly optimizes economic, environmental, service, resilience, energy, and operational-risk objectives under fuzzy uncertainty. Evaluation was conducted using combined real and statistically consistent simulated data across six operational scenarios ranging from baseline conditions to a critical scenario involving simultaneous demand growth, capacity restrictions, cost escalation, uncertainty, and stakeholder conflicts. Results demonstrate progressive improvements in decision quality under increasingly complex conditions; in the critical scenario, Human–AI collaboration achieved a 19.6% cost improvement, while the service level reached 99.4%. The findings demonstrate that integrating cognitive representation, collaborative intelligence, strategic adaptation, and fuzzy optimization provides a unified mechanism for adaptive multi-stakeholder supply-chain decision-making.