Amid increasing exposure to supply chain (SC) risks rooted in environmental, social, and governance (ESG) conformity requirements, this research presents a three-stage confirmatory factor analysis (CFA) that validates ESG risk (ESG-R) as a coherent second-order construct. A structural equation model (SEM) shows that higher ESG-R significantly increases Mineral SC vulnerability. A regional assessment further highlights substantial variation in ESG across Africa, the Americas, Asia, Europe, and Oceania, containing both extremely high-risk and low-risk countries, demonstrating that geography alone does not determine institutional or environmental fragility. Principal component analysis identifies two structural dimensions: cumulative ESG-R and environmental imbalance, forming the basis for a four-quadrant typology of country risk profiles. Cluster analysis demonstrates that high-risk countries differ not only in magnitude but in the composition of their vulnerabilities. The research provides a robust, data-driven framework for identifying high-risk mineral supply regions and offers actionable insights for policymakers and firms seeking to strengthen responsible sourcing and enhance SC resilience.

Urban change is intensifying the interdependencies among infrastructure systems, governance processes, and social dynamics, particularly in fast-growing cities exposed to compound risks and rapid transformation. These conditions challenge existing community resilience assessment approaches, which often rely on static indicators and standardized measurement structures that insufficiently capture contextual variability and evolving community capacities. This paper reviews community resilience assessment research to identify methodological gaps and emerging trends. A systematic approach is used to analyze peer-reviewed literature published between 2016 and 2026. The review highlights a dominant strand of resilience scholarship centered on community-focused approaches while revealing fragmentation in how assessment structures are designed and applied. Findings show that many approaches rely on static indicator sets, weakly integrate governance processes, and insufficiently incorporate community participation and social dynamics. The analysis identifies critical limitations in community resilience assessment, including fragmented data environments, limited local-scale calibration, and inconsistent measurement logic across studies. Building on this analysis, the paper introduces a diagnostic interpretation that explains fragmentation as misalignment across hazard orientation, measurement architecture, governance capacity, and socio-economic calibration. To address these gaps, the paper develops a conceptual diagram that synthesizes how key components of community resilience assessment interact in fast-growing urban contexts. The study argues that advancing community resilience requires approaches that combine data-informed insights with participatory governance, enabling more adaptive, locally grounded, and operationally meaningful understanding of community resilience processes.

This paper presents a comparative evaluation of six traffic state estimation methods applied to an urban road network modelled as a Cyber-Physical System subject to partial observability induced by sparse probe vehicle data. Four established baseline methods are considered: Exponential Moving Average (EMA), Adaptive EMA (AEMA), one-dimensional Kalman Filter, and Blended Meta-Estimator (BME). Two original contributions are introduced: the Weighted Observation Smoother (WOS) and the Regime-Switching Dual EMA (RSDE). The experimental framework consists of a synthetic urban network of 120 road segments simulated over 720 time steps at a 30-second resolution, under connected vehicle penetration rates of 5%, 10%, and 20%, incorporating a non-recurrent incident event with explicit spatial propagation. Performance is assessed in terms of root mean square error, mean absolute error, median incident detection delay, and multi-horizon prediction error at horizons of one, three, and five steps. Results indicate that WOS achieves the lowest RMSE at penetration rates of 10% and 20%, while RSDE demonstrates the best incident detection latency at 10% penetration and competitive multi-horizon accuracy at a forecast horizon of five steps. At 5% penetration, classical filters remain competitive, suggesting that method selection in real-world deployments should be conditioned on the estimated in-field penetration rate.

The Product Environmental Footprint (PEF) methodology is the European Commission’s reference framework for substantiating environmental claims, yet its uptake remains limited, especially among SMEs and sectors with complex global value chains. A key scalability barrier is the predominantly product-specific development of Product Environmental Footprint Category Rules (PEFCRs), which often requires substantial modelling effort, expert knowledge, and data availability for each new product group. This paper proposes a conceptual framework for cross-sectoral PEFCR development based on modularity and mirroring, i.e., the controlled reuse of validated life-cycle modules across categories under explicit equivalence criteria. The framework defines modular PEF components and mirroring nodes and introduces an inheritance-tree logic to support systematic reuse and updateability of verified modules across related product categories. The approach is illustrated through home textiles and detergents, highlighting shared hotspots in the use phase as promising candidates for module reuse. By structuring PEFCRs around reusable, governed building blocks, the framework aims to reduce duplicated modelling work, improve methodological consistency, and better align PEFCR design with emerging digital requirements (e.g., Digital Product Passports) and regulatory needs.