AI-Driven Digital Product Passports for Sustainable Textile Supply Chains

Abstract

The textile industry faces increasing pressure to improve supply chain transparency, environmental sustainability, and compliance with emerging circular economy regulations. Although Digital Product Passports (DPPs) have been recognized as a promising mechanism for capturing product lifecycle information, existing implementations primarily focus on static traceability and regulatory documentation, offering limited capabilities for intelligent sustainability assessment, predictive analytics, and automated decision support. Furthermore, current studies often investigate blockchain, Internet of Things (IoT), or artificial intelligence (AI) as isolated technologies, leaving a significant research gap in developing an integrated framework that combines secure product traceability with AI-driven sustainability intelligence for textile supply chains. To address this limitation, this paper proposes an Artificial Intelligence-Driven Digital Product Passport (AI-DPP) framework that integrates machine learning, blockchain, IoT-enabled data acquisition, and lifecycle sustainability analytics into a unified architecture. The proposed framework generates comprehensive digital passports containing verified information on material composition, manufacturing processes, environmental indicators, transportation history, certifications, and recycling attributes while employing AI models to predict overall sustainability performance, supplier risk, carbon efficiency, water sustainability, and recyclability. Blockchain technology ensures secure, immutable, and transparent management of passport data across all supply chain participants. The framework is evaluated using a simulated dataset comprising approximately 50,000 textile product records developed from publicly available sustainability datasets. Comparative analysis of multiple machine learning algorithms demonstrates that XGBoost achieves the highest predictive performance with an R² score of 0.947, while the blockchain layer maintains 100% data integrity and a 99.82% transaction verification success rate. Experimental results further indicate significant improvements in product traceability, supplier visibility, recycling recommendation accuracy, and environmental sustainability metrics compared with conventional traceability approaches. The proposed AI-DPP framework offers a scalable and intelligent solution for supporting circular textile supply chains by enabling predictive sustainability management, trustworthy product lifecycle transparency, and data-driven decision-making for manufacturers, regulators, consumers, and recycling stakeholders.