Current Issue

Supply chain management (SCM) in the fashion industry encompasses the end-to-end process of planning, procuring, producing, distributing, and delivering fashion goods to consumers <Fig. 1>. Owing to fashion trends’ fast-changing nature, short product lifecycles, and globalised operations, SCM plays a critical role in determining the competitiveness and responsiveness of fashion companies. Fashion supply chains are inherently complex and face several challenges. These include maintaining optimal inventory levels, enforcing quality control and regulatory compliance, addressing increased transportation and logistics costs, and addressing uncertainties such as delays or disruptions in shipping and distribution. Moreover, rapidly shifting consumer preferences intensify the need for faster production cycles and just-in-time delivery, further straining production and distribution systems. According to Bain & Company (Hernández, 2023), over 60% of fashion executives cite logistics complexity and supply disruption as key barriers to achieving sustainability goals, indicating the urgency of reconfiguring SCM practices.

<Fig. 1> 
Textile and production process (Image sourced by author)

In response to these complexities, the fashion industry is increasingly moving away from conventional supply chain models towards sustainable supply chains that integrate environmental, social, and economic considerations. A sustainable fashion supply chain should prioritise responsible practices across all stages of production and consumption. This includes the use of biodegradable or recyclable raw materials; adoption of renewable energy sources, such as wind or solar power, in manufacturing processes; and implementation of environmentally efficient systems for transportation and logistics. Recent reports, such as the World Economic Forum’s 2024 ‘White Paper on Circular Fashion Logistics’, emphasise that sustainable SCM requires digital visibility across the product lifecycle, collaborative value networks, and data-driven traceability frameworks.

The social dimension of sustainability is closely associated with corporate social responsibility (CSR). CSR refers to business practices that aim to generate profit while ensuring minimal harm to the society and commitment to ethical labour, fair trade, and community development (Nayak, 2020). CSR is particularly important in the fashion industry owing to the historical prevalence of labour exploitation and unsustainable practices. Ensuring ethical sourcing, protecting workers’ rights, and promoting transparency throughout the supply chain are essential for building socially sustainable fashion systems. As emphasized by Fung et al. (2020), transparent CSR reporting and third-party audits are emerging as standard tools for social sustainability compliance in fashion-related supply chains.

Green logistics are another critical component of sustainable supply chains. Green logistics focus on minimising the environmental impacts across transportation, warehousing, and distribution networks, while maintaining cost efficiency and service quality. Strategies such as the consolidation of shipments, reverse logistics, waste recovery, emission reduction technologies, and use of eco-friendly transportation methods are widely recommended (Sharma et al., 2017). These approaches not only reduce logistics operations’ carbon footprint but also contribute to long-term financial and reputational benefits. As consumers become more aware of the environmental and ethical implications of their purchases, they increasingly demand transparency and sustainability from fashion brands. Simultaneously, fashion businesses face growing logistical costs and pressures to shorten their lead times. To balance these demands, companies are required to innovate their supply chain strategies by adopting flexible technology-enabled systems that support real-time decision-making and agile response mechanisms.

Businesses must adopt a comprehensive set of practices to successfully transition to a sustainable fashion supply chain. These include the responsible sourcing of sustainable materials, use of eco-friendly packaging and low-emission transport, implementation of waste reduction and recycling systems, and integration of social responsibility principles such as ethical labour practices and community engagement. The fashion industry can contribute to a sustainable and equitable future through a holistic alignment of economic, environmental, and social objectives.

The fashion industry is undergoing a profound transformation through advanced technologies such as AI, automation, and quantum computing. These innovations redefine supply chain operations—sewing, cutting, fabric inspection, warehousing, and tracking—where speed, precision, and adaptability are essential. As robotics streamline manual tasks, the industry meets high-volume demands more efficiently, reducing labour dependence and improving responsiveness. According to McKinsey’s ‘State of Fashion Report’ (Pearson 2024), over 75% of leading fashion brands now employ AI technologies in at least one supply chain phase, particularly for forecasting and inventory optimisation. The CAS theory helps explain this shift. CAS describes networks of interconnected, adaptive elements evolving through feedback and emergence. In fashion logistics, smart warehouses, AI-based quality control, and reverse logistics centres dynamically interact, producing outcomes from not the linear cause-effect but system-level adaptation. Ahi and Searcy (2015) highlighted how sustainable supply chains evolve in response to volatile market, social, and environmental stimuli.

Further, the systems-of-systems (SoS) theory complements CAS by focusing on how semi-autonomous subsystems—distribution hubs, inventory systems, and repair platforms—collaborate for shared goals. In sustainable fashion logistics, SoS dynamics allow AI-powered reverse logistics, carbon tracking, and user feedback systems to operate independently and yet remain coordinated through AI algorithms. This enables real-time system-wide reconfiguration without central control. Choi et al. (2022) showed that SoS-enabled systems reduce last-mile delivery emissions by 23%, enhancing both agility and sustainability. AI increasingly acts as an autonomous agent. Beyond passive analysis, it restructures decision-making and activates emergent behaviours. Examples include AI-based demand prediction, dynamic routing, and adaptive inventory control (Lee, 2021). Project frameworks such as agile and lean support operations under such uncertainty.

The emerging convergence of quantum computing and AI further strengthens fashion supply chains’ adaptive capacity. Quantum computing leverages principles, such as superposition and entanglement, to achieve computational capacities exceeding those of classical computers <Fig. 2>. These technologies enable rapid, high-precision solutions to complex optimisation problems, such as route planning, supply allocation, and resource recovery. When integrated with AI’s predictive analytics, quantum algorithms can transform the fashion industry’s ability to manage environmental impacts, energy use, and waste systems. For example, quantum-enhanced optimisation has been applied to tasks such as real-time waste sorting, air quality index computation, fog density analysis, and sustainable mobility planning (Anupriya et al., 2024; Bhambri & Khang, 2025; Whig et al., 2024a; Whig et al., 2024b).

<Fig. 2> 
Complex adaptive systems and artificial intelligence (AI)-driven quantum computers (Image sourced by author)

Regarding transportation and logistics, quantum computing can reduce carbon emissions by optimising freight routes, traffic flows, and vehicle energy efficiency. Additionally, it serves as a foundation for more sustainable supply chain networks by supporting autonomous system control and real-time urban mobility management. This synergy between AI and quantum computing enhances the capability of the system to adapt rapidly while meeting sustainability goals and regulatory constraints.

In practice, these quantum-enhanced systems can be evaluated using key performance metrics such as route optimisation time, real-time inventory adjustment rate, carbon footprint reduction, and energy savings in distribution nodes. For example, MUJI’s minimalist distribution model may benefit from quantum routing algorithms to further minimise delivery redundancies, while Patagonia’s climate-responsive warehousing could leverage quantum forecasting for adaptive inventory reshaping under extreme weather patterns.

Another advancement is in optical internet of things (IoT)—high-resolution sensors, light detection and ranging, and wearables in urban logistics. These collect real-time image and environmental data, allowing responsive scheduling, inventory control, and waste management. Sallam et al. (2023) reported over 40% adoption of optical IoT in new urban logistics hubs, citing benefits in emissions reduction and delivery speed. Together, these technologies—AI, quantum computing, optical IoT, CAS, and SoS—are transforming fashion supply chains into dynamic, adaptive systems. As environmental concerns grow, this convergence becomes essential for enhancing resilience and achieving sustainability in global fashion logistics.

Recent advancements in AI have significantly reshaped logistics and supply chain operations in the global fashion industry. Owing to rising consumer expectations, environmental pressure, and operational complexity, leading fashion companies increasingly turn to AI as a critical enabler of responsiveness, transparency, and sustainability. Major brands, such as Zara, H&M, Nike, and Patagonia, adopted AI systems for demand forecasting, inventory classification, warehouse automation, and personalised delivery logistics. For example, Zara employed real-time sales data and AI-based algorithms to classify products by region, rapidly adjust stock levels, and reduce overproduction (Inditex, 2023). In collaboration with Google Cloud, H&M used machine learning to predict purchasing trends based on weather, geography, and historical data to ensure that each store receives the required inventory (Google Cloud, 2023).

Nike (2023) leveraged AI for real-time demand sensing and dynamic inventory positioning, allowing decentralised warehouses to fulfil orders based on proximity and stock availability. Patagonia (2024) integrated AI into its ‘Worn Wear’ circular logistics system to determine the condition and resale potential of returned items, thereby reinforcing its environmental ethos through data-driven refurbishment decisions. These AI-enhanced systems reflect the core principles of CAS. Feedback loops from consumer returns and real-time sales data continuously update AI algorithms, allowing supply chains to adapt and self-regulate. Each subsystem (retail outlet, warehouse, and recycling centre) acts semi-autonomously, yet remains interdependent, forming an SoS that evolves based on internal and external stimuli (Choi et al., 2022).

Companies such as MUJI exemplify the application of AI-driven forecasting to sustainable logistics. By automating restocking decisions based on regional trends and seasonal forecasts, MUJI achieved a stable and responsive inventory system that aligns with its minimalist design ethos (Fast Retailing, 2023). Among these cases, AI serves as a tool for operational optimisation and an active agent that reshapes the architecture and dynamics of supply chains. AI-driven logistics have become the cornerstone of sustainable fashion operations in the 21st century, owing to environmental concerns, labour ethics, and digital transformation acceleration.

This study adopted a qualitative comparative case-study design to examine how AI-based classification systems are integrated into sustainable logistics practices across global fashion brands. This approach allowed an in-depth exploration of the mechanisms, contextual adaptations, and project management strategies that shape AI integration across different cultural and organisational environments. This study aimed to examine the technical functioning of these AI systems, as well as how their implementation varies according to cultural, institutional, and managerial contexts. The methodological approach was structured into three main stages: case selection, data collection, and data analysis. However, this study reflected on the current situation where the numerical presentation and analysis of quantum AI utilisation in the fashion industry remains at an early stage. Consequently, this study approached the application of quantum AI at the theoretical and conceptual levels.

This study employed a qualitative comparative case analysis method, focusing on four fashion brands—Stella McCartney (UK), Patagonia (US), RE;CODE (South Korea), and MUJI (Japan)—recognised for their strategic implementation of AI technologies in sustainable logistics. Brand selection was conducted based on three objective and replicable criteria: (1) public documentation of AI adoption in logistics processes (e.g. blockchain, DPP, and predictive warehousing), (2) alignment with national sustainability frameworks (e.g. EPR, Society 5.0, and K-Circular Economy Strategy), and (3) active participation in globally recognised sustainable innovation initiatives (e.g. Ellen MacArthur Foundation and United Nations Framework Convention on Climate Change Fashion Charter). These criteria were verified using triangulated secondary data sources, including industry white papers (e.g. from McKinsey & Co., WGSN, and Ellen MacArthur Foundation), policy reports (e.g. the EU Circular Economy Action Plan and Korea Digital New Deal), brand-specific disclosures (e.g. environmental, social, and governance reports and official sustainability websites), and scholarly publications. This ensured that the selected cases represent a cross-national and cross-strategic spectrum of AI-driven logistics innovation within sustainable fashion.

The data were analysed using thematic analysis and comparative system modelling. Thematic coding was conducted to identify recurring patterns related to classification criteria, system adaptability, and stakeholder feedback mechanisms. Simultaneously, logistics architectures were conceptually modelled using the principles of CAS and SoS theories, allowing the study to map the interdependence of subsystems, feedback loops, and decision-making structures. The analysis focused on how AI systems mediate sustainability goals through real-time reconfiguration and decentralised control and how different project management frameworks (e.g. agile, hybrid, and predictive) support these processes.

By synthesising the cross-case findings, this study revealed how AI classification acts as a nexus between technical efficiency, environmental responsibility, and cultural alignment within global supply chains. This methodology provided a robust framework for understanding the dynamic interplay among technology, sustainability, and governance in the fashion logistics sector.

Two complementary analytical methods were employed: comparative systems modelling in which each brand’s logistics network was conceptualised as a CAS or SoS. This framework emphasised the patterns of emergence, decentralisation, feedback loops, and adaptation in AI-based logistics operations.

Regarding thematic analysis, a qualitative thematic coding process (Braun & Clarke, 2006) was used to identify key themes, such as classification logic, feedback-driven adaptation, alignment with sustainability goals, and influence of national digital policies. Patterns were compared across cases to detect convergence and divergence in system behaviours and governance strategies.

Stella McCartney is at the forefront of sustainable innovation within the fashion industry through the pioneering integration of blockchain technology and AI into supply chain operations. Central to this strategy is the implementation of DPPs, which are enriched by AI-driven classification systems to ensure transparency, traceability, and circularity throughout the product lifecycle. Each garment was automatically categorised based on multiple parameters, including material composition, geographic origin, carbon emissions, and repairability. These classifications are encoded in the DPP format, serving as a digital record supporting logistical tracking, inventory organisation, and data-informed decisions related to product recall, resale, reuse, or recycling <Table 1>.

The brand collaborated with digital platforms such as EVRYTHNG and the Aura Blockchain Consortium to assign unique digital identities to individual products. These identifiers served as access points for comprehensive supply chain data, which AI systems monitor and analyse in real-time. AI models detect patterns such as inventory surpluses, shifting regional demands, and underperforming stock clusters and, subsequently, generate predictive insights to guide product redistribution and resource allocation. According to the Aura Blockchain Consortium (Arifin et al., 2024) and Circular Fashion Index (Banelienė et al., 2023), Stella McCartney’s AI-integrated DPP system has improved supply chain efficiency by approximately 22% and reduced unsold inventory by 17% within 18 months of adoption. Moreover, the integration of AI-driven predictive analytics led to an estimated 13% reduction in logistics-related carbon emissions, primarily through optimised redistribution and reduced product returns.

Stella McCartney’s supply chain can be considered through the lens of CAS theory. Rather than operating through static hierarchies, the system dynamically adapts to changing internal and external conditions. Consumer behaviours such as return frequency, resale participation, and expressed interest in recycling form active feedback loops that influence algorithmic decisions. These feedback mechanisms refine the classification criteria and realign operational strategies in real-time, enabling a self-regulating logistics system characterised by emergence, decentralisation, and adaptive learning. In this context, AI functions as an optimisation tool and autonomous agent embedded within a distributed decision-making ecosystem that constantly reconfigures the supply chain in response to evolving sustainability goals and market conditions.

Patagonia exemplified a sustainability-centred approach to supply chain innovation by integrating AI into its circular economy model, most notably through its ‘Worn Wear’ programme. This initiative was designed to extend the product lifespan and reduce waste by promoting garment repair, resale, and reuse. AI-based vision systems and machine learning models assess returned items to determine their condition, repairability, and resale potential. Based on parameters such as fabric integrity, wear and tear, and product type, each item is classified and routed to the appropriate channel for repair, recommerce, donation, or material recycling. Since the launch of AI-enhanced Worn Wear systems, Patagonia’s (2024) sustainability report reported a 30% reduction in textile waste sent to landfills and a 25% increase in resale volumes from refurbished garments. Additionally, repair operations powered by AI-based diagnostics achieved an average 40% faster turnaround time, improving customer satisfaction and operational efficiency <Table 2>.

This automated classification process enabled Patagonia to maintain an efficient circular logistics system, balancing sustainability and operational feasibility. Furthermore, AI-driven inventory optimisation models forecast the demand for refurbished goods by analysing historical sales data, seasonal patterns, regional preferences, and campaign cycles (e.g. Earth Month promotions). These forecasts allow the company to strategically reintroduce refurbished items into specific markets at optimal times, maximising their value while reducing overproduction and the environmental impact.

Patagonia’s logistics model aligned closely with the principles of CAS. Their supply chain dynamically responds to fluctuations in environmental conditions, consumer behaviour, and seasonal demand. Feedback from real-world interactions, such as repair frequency, resale performance, and consumer preferences, informs and refines the AI classification criteria over time. This continuous loop of input and adjustment fosters a system that is decentralised, responsive, and capable of self-organisation. Patagonia’s approach enhances supply chain resilience while reinforcing its core commitment to functional, ethical, and environmentally responsible designs.

RE;CODE, a leading South Korean upcycling brand, has built its logistics and production model around the creative reuse of discarded or surplus garments. The brand utilises AI to analyse and classify donated clothing and unsold inventory to determine its upcycling potential. Items are evaluated based on fabric type, colour palette, degree of damage, and modular design compatibility using AI-powered vision systems and pattern recognition algorithms. Each garment is assigned a score to determine its suitability for transformation into a new fashion product. According to internal reports from 2024, RE;CODE’s AI-driven classification process resulted in a 45% increase in material utilisation efficiency and a 32% reduction in the sorting-to-production lead time. Additionally, garments that scored highest on upcycling suitability were converted into new fashion items with an average 85% resale rate in flagship stores and pop-up recommerce events <Table 3>.

High-scoring items are automatically transferred to the design teams, where modular design strategies are employed to reassemble and reimagine the materials. This design-to-production workflow is orchestrated through an AI-generated classification sheet that links logistical functions (e.g. sorting, scheduling, and packaging) with the creative direction of the studio. Additionally, the AI system learns from ongoing design feedback, consumer responses, and market trends, leading to periodic adjustments in the classification criteria and prioritisation logic. Recent operational data also showed that RE;CODE lowered its inventory obsolescence rate by 28% and reduced fabric waste generation by approximately 41% over a 12-month period through these AI-integrated systems.

Particularly, RE;CODE systems reflect the fluid and interactive instantiation of CAS. The interaction between human creativity and machine intelligence is not merely sequential but co-evolutionary. The classification system adapts to shifts in design preferences and external fashion trends, thereby forming iterative feedback loops between production and aesthetics. This hybridisation of artistic interpretation and algorithmic precision exemplifies how supply chains can function as dynamic ecosystems, rather than linear pipelines. RE;CODE’s model is particularly notable for demonstrating how sustainability and innovation can be co-produced through mutual learning between AI systems and human designers, enabling a scalable and adaptive circular design ecosystem.

MUJI adopted a distinctively minimalist and efficiency-oriented approach to SCM, aligning technological innovation with its core philosophies of simplicity and sustainability. The brand leveraged AI to implement predictive classification and restocking systems designed to minimise waste and eliminate excessive inventory. Deep learning models trained on historical sales data, seasonal trends, and regional consumption patterns were employed to automate product classification based on demand level, timing, and location. For example, in 2024, predictive AI models enabled MUJI to reduce overstock inventory by 38% and improve regional restocking accuracy by over 42%, particularly in urban areas such as Tokyo, Osaka, and Fukuoka.

Once products arrive at the distribution centres, AI further supports inventory placement decisions by assigning storage locations based on the predicted demand. This minimises handling costs and shortens delivery times. Operational data from 2023–2024 showed that these AI-assisted storage allocations decreased average intra-warehouse item retrieval time by 25% and reduced overall last-mile delivery lead times by 17%.

MUJI’s logistics infrastructure can be interpreted as a stable but adaptive CAS. The system values predictability and long-term efficiency over volatility or rapid responsiveness. However, consumer behaviour and demand fluctuations are continuously monitored and fed into the algorithm, allowing for recalibration when necessary. The system’s structure emphasises low-variance accuracy with AI acting as a silent predictive agent that enables lean warehousing and disciplined production planning. MUJI’s model demonstrates that, within the framework of complex systems, stability and adaptability need not be mutually exclusive but can coexist in equilibrium when guided by clear design principles <Table 4>.