Abstract:

Ground‑mounted photovoltaics, including agrivoltaic concepts, are increasingly deployed on agricultural land. In practice, damaged modules from repowering modules are sometimes stored on‑site for prolonged periods, creating localized vegetation suppression and land‑stewardship concerns that are rarely quantified. We present two anonymized case studies from Czechia (nominal capacities of 0.861 and 1.109 MWp; commissioned 2010 and 2009; repowered 2022 and 2021), where cracked backsheets and/or broken front‑glass modules were stacked and stored directly on grasslands within PV parcels. Using GIS delineation on orthophotos supported by field photographs, we quantified the land area (19,560 and 22,100 m²), PV panel area (plan‑ view; 4,960 and 5,080 m²), and stored PV module area (plan‑ view storage footprint; 109 and 100 m²). Stored module counts were estimated from visible stacks (≈1800 and ≈2000 modules). Using a conservative mass range of 18–25 kg/module, the stored masses were ~32–45 t and ~36–50 t, respectively. Although the storage footprints constitute <1% of the land area, they create persistent “dead zones” on agricultural land and concentrate tens of tonnes of material directly on the soil. We discuss regulatory and economic barriers to timely removal in the context of circular‑economic goals and propose practical reporting indicators for repowering projects on agricultural land: A_store (m²), N_store (pcs), M_store (t), storage duration, condition class, and storage interface.

Abstract: High-dimensional survival analyses require calibrated risk and honest uncertainty, but standard elastic-net Cox models yield only point estimates. We develop a fully Bayesian elastic-net Cox (BEN–Cox) modelfor high-dimensional proportional hazards regression that places a hierarchical global–local shrinkage prior on coefficients and performs full Bayesian inference via Hamiltonian Monte Carlo. We represent the elastic–net penalty as a global–local Gaussian scale mixture with hyperpriors that learn the ℓ1/ℓ2 trade-off, enabling adaptive sparsity that preserves correlated gene groups and, using HMC on the Cox partial likelihood, yields full posteriors for hazard ratios and patient-level survival curves. Methodologically, we formalize a Bayesian analogue of the elastic-net grouping effect at the posterior mode and establish posterior contraction under sparsity for the Cox partial likelihood, supporting the stability of the resulting risk scores. On the METABRIC breast-cancer cohort (n = 1 , 903; 440 gene-level features from an Illumina array with ≈ 24,000 gene-level features (probes)), BEN–Cox achieves slightly lower prediction error, higher discrimination, and better global calibration than a tuned ridge Cox baseline on a held-out test set. Posterior summaries provide credible intervals for hazard ratios, identify a compact gene panel that remains biologically plausible. BEN–Cox provides a theory-backed, uncertainty-aware alternative to tuned penalised Cox models, improving calibration and yielding an interpretable sparse signature in correlated, high-dimensional survival data

Abstract: This paper seeks to analyze the philosophy of marriage in India as a construct based on three distinct and conflicting models: the contract, the institution, and the moral bond. The primary focus is to consider how the marriage contract, as a sacred Muslim Nikah and a secular civil agreement under the Special Marriage Act, 1954 and Hinduism as a sacred event and also, a civil agreement under the Hindu Marriage Act, 1955. The moral bond is represented by widows and modern-day “companionate” partnership. It concludes that Indian marriage is a struggle between all three models due to globalization, post-colonial feminist critiques of its patriarchal nature, and the individualization of Western ideals around partnership and friendship. The quintessential example of all three struggles is love-cum-arranged marriage.

Abstract: Alcohol-associated liver disease (ALD) is driven by complex interactions among hepatic lipid accumulation, oxidative stress, inflammation, cell death, and disruption of the gut–liver axis. Therapeutic strategies capable of targeting multiple interconnected pathogenic pathways remain limited. In this study, we investigated the protective potential of graphene oxide nanoparticles (GNPs) in a chronic ethanol-fed rat model of ALD. Male Wistar rats were subjected to ethanol feeding and intermittently treated with GNPs (10 mg/kg) by oral gavage. Hepatic injury was assessed by biochemical parameters, histology, lipid accumulation, gene and miRNA expression, protein analysis, and gut microbiome profiling. Ethanol feeding induced hepatic steatosis, oxidative stress, apoptotic and necroptotic signaling, intestinal barrier disruption, gut dysbiosis, and activation of hepatic inflammatory pathways. GNP treatment markedly attenuated ethanol-induced lipid accumulation, normalized liver morphology, and reduced biochemical markers of liver injury. These effects were accompanied by restoration of antioxidant defenses, including Nrf2 and HO-1, and suppression of CYP2E1 expression and cell death–associated markers. In parallel, GNPs preserved intestinal architecture, maintained tight junction gene expression, and suppressed intestinal inflammatory responses. Gut microbiome analysis revealed partial restoration of ethanol-induced dysbiosis, including recovery of beneficial postbiotic-associated bacterial taxa. Improved intestinal homeostasis was associated with attenuation of hepatic TLR4-associated inflammatory signalling and modulation of macrophage-associated markers. Furthermore, GNP treatment partially normalized ethanol-induced dysregulation of miRNAs implicated in lipid metabolism, inflammation, and oxidative stress. Collectively, these findings demonstrate that GNPs exert a coordinated protective effect against ethanol-induced liver injury by modulating multiple pathological processes along the gut–liver axis. This multi-targeted activity highlights the therapeutic potential of graphene oxide nanoparticles as an intervention strategy for early-stage alcohol-associated liver disease.

Abstract: Real-world data streams, such as news articles and social media posts, are dynamic and nonstationary, creating challenges for real-time structured representation via knowledge graphs, where relation extraction is a key component. Continual relation extraction (CRE) addresses this setting by incrementally learning new relations while preserving previously acquired knowledge. This work investigates the use of pretrained language models for CRE, focusing on large language models (LLMs) and the effectiveness of memory replay in mitigating forgetting. We evaluated decoder-only models and an encoder-decoder model on TACRED and FewRel in English. Our results show that memory replay is most beneficial for smaller instruction-tuned models (e.g., Flan-T5 Base) and base models such as Llama2-7B-hf. In contrast, the remaining instruction-tuned models in this work do not benefit from memory replay, yet some, like Mistral-7B, already achieve higher accuracies without it and surpass prior methods. We further observed that Llama models in this work are more prone to hallucinations. To the best of our knowledge, this work provides the first reproducible benchmarks for LLMs in CRE. It offers a novel analysis of knowledge retention and hallucination behavior—dimensions that have not been systematically studied in earlier research.

Abstract: Sentiment classification struggles with complex semantic relationships using static text graphs. We introduce the Quantum-Enhanced Adaptive Graph Convolutional Network (QAGCN), a hybrid quantum-classical architecture for robust sentiment representation. QAGCN's core is a Quantum-Enhanced Graph Construction Module employing a Parameterized Quantum Circuit (PQC) to dynamically learn emotional association strengths between word pairs. This generates a task-adaptive adjacency matrix, which then feeds into classical GNN layers. Evaluations on benchmark datasets (Yelp, IMDB, Amazon, MC, RP) demonstrate QAGCN's superior or competitive accuracy against state-of-the-art classical graph models and the Quantum Graph Transformer. QAGCN notably improved performance on Amazon where prior quantum models struggled, underscoring its adaptive graph construction's efficacy. An ablation study confirms the critical contribution of PQC-driven adaptive graph learning. Our findings highlight the significant potential of quantum-enhanced adaptive graph learning for complex Natural Language Processing.

Abstract: E-Payment has become popular in mobile com-merce, can provide consumers with a convenient way to makepurchases electronically. Currently, however, all too many E-Payment systems are primarily focused on securing a consumer’sfinancial information and do little to prevent privacy leaks andAI-generated scams. This paper defines AEP-M, a novel AI-enhanced anonymous e-payment scheme developed for mobiledevices that uses TrustZone and divisible e-cash. Since mobiledevices have very limited processing power and each transactionmust be performed in real time, the proposed solution combinesan efficient divisible e-cash system with AI-powered anomalydetection techniques to improve both the security, privacy andfraud detection in mobile payments. In addition to enablingusers to divide a single withdrawal of an e-coin of a largeamount into multiple transactions without disclosing their iden-tity to either banks or merchants, AEP-M integrates AI-basedrisk assessment to identify suspicious spending behaviors torapidly mitigate fraud and continuously monitor transactions.By employing a combination of bit decomposition and pre-computation to minimize the computational overhead of thetransaction process, AEP-M provides the optimal performancein terms of minimizing the max number of exponentiationoperations required to perform the frequent online spendingprocess on elliptic curves. Finally, AEP-M also incorporates anARM TrustZone to protect a user’s financial data and importantprivate data; an SRAM PUF is used as a Root of Trust to deriveAI-powered keys and manage sensitive data, thereby increasingboth the security and reliability of the system. A prototype ofAEP-M was implemented and evaluated using the BN curve ata 128-bit security level. The experimental results demonstratedthat AEP-M is capable of improving the Security, Efficiency andFraud Detection capabilities of Mobile Digital Payments whilemaintaining User Privacy and Anonymity.

Abstract:

The selection of an optimal antifoam is critical for efficient fermentation, as industrial agents often have detrimental side effects like growth inhibition, while some can enhance productivity. This study presents a rational approach to developing and screening novel silicone-polyol antifoam emulsions. A key finding was the discovery of selective antibacterial activity in agent 3L10, which strongly inhibited Gram-positive bacteria (especially Corynebacterium glutamicum) but not Gram-negative strains. This specificity, likely mediated by interaction with the mycolic acid layer of C. glutamicum, highlights the necessity for strain-specific antifoam testing. A comprehensive evaluation protocol—combining chemical design, cytotoxicity screening across diverse microorganisms, determination of minimum effective concentrations (MEC), and validation in model bioreactor fermentations—was established. Through this process, agent 6T80 was identified as a promising candidate. It exhibited low MEC, high emulsion stability, no cytotoxicity, and did not impair growth or recombinant protein production in B. subtilis or P. putida fermentations. The study concludes that agent 6T80 is suitable for further application in processes involving Gram-negative and certain Gram-positive hosts, whereas agent 3L10 serves as a valuable tool for studying surfactant-membrane interactions. The developed methodology enables the targeted selection of highly efficient and biocompatible antifoams for specific biotechnological processes.

Abstract: Precise quantification of fine motor behavior is essential for understanding neural circuit function and evaluating therapeutic interventions in neurological disorders. While markerless pose estimation frameworks such as DeepLabCut (DLC) have transformed behavioral phenotyping, the choice of convolutional neural network (CNN) backbone significantly impacts tracking performance, particularly for tasks involving small distal joints and partial occlusions. in this paper, we present the first systematic comparison of nine CNN architectures implemented in DLC for lateral-view analysis of fine reaching movements in the Montoya Staircase test, a gold standard assay for skilled forelimb co-ordination in rodent models of stroke and neurodegenerative disease. Using a dataset of videos representing both control and M1-lesioned conditions, we rigorously evaluated models across six critical dimensions: spatial accuracy (RMSE, PCK@5px), mean average precision (mAP), occlusion robustness, inference speed and GPU memory usage. Our results reveal that multi-scale DLCRNet architectures substantially outperformed classical backbones, with DLCRNet_ms5 achieving the highest overall accuracy and DLCRNet_stride16_ms5 providing the best trade-off between precision and computational efficiency. These findings provide critical methodological guidance for neuroscience la-boratories and highlight the importance of architecture selection for rigorous quantification of fine motor behavior in preclinical research.

Abstract: Accurate detection of buried utilities and reliable characterization of shallow subsurface conditions are critical requirements in civil and industrial engineering projects, particularly in urban areas developed over conductive clay–marl formations. In such environments, commonly used electromagnetic techniques often fail due to severe signal attenuation, increasing uncertainty during excavation and infrastructure planning. This study presents a high-resolution engineering workflow based on Electrical Resistivity Tomography (ERT) for the simultaneous detection of buried stormwater and sewer pipes and the geotechnical characterization of shallow subsurface materials. The methodology was applied in an industrial area southwest of Pamplona (Navarra, Spain), where Eocene marls and clays dominate the geological setting. Three ERT pro-files, each 23.5 m long, were acquired using a pole–dipole array with a dense electrode spacing of 0.5 m, allowing decimetric-scale resolution and investigation depths of up to 7–8 m. Data were processed and inverted using both smooth (L2-norm) and robust (L1-norm) inversion schemes to evaluate their influence on anomaly detection and stratigraphic imaging. The resulting resistivity models clearly identified elongated conductive and resistive anomalies corresponding to known buried sewer and stormwater pipes, despite the highly conductive background. In addition, the ERT sections revealed lateral and vertical variations within the clay–marl sequence, including sandy and compact detrital facies of direct relevance for foundation design and excavation planning. Borehole data available in the study area corroborated the geophysical interpretation. A complementary Ground Penetrating Radar (GPR) survey confirmed the ineffectiveness of electromagnetic methods under the same conditions due to rapid signal attenuation. Rather than focusing solely on utility detection, the proposed approach frames ERT as a dual-purpose engineering tool capable of providing continuous subsurface infor-mation that bridges the gap between sparse borehole data and construction needs. The workflow presented here is transferable and scalable, offering a practical protocol for urban and industrial projects in conductive soils where conventional techniques are limited.

Abstract: A type II cadherin, Cadherin-19 (CDH19), plays a vital role in neural crest development. CDH19 regulates cell–cell junctions and migration by forming catenin-cytoskeleton complexes. Although anti-CDH19 monoclonal antibodies (mAbs) are used for specific applications such as Western blotting and immunohistochemistry (IHC), suitable anti-CDH19 mAbs for flow cytometry are limited. Here, novel anti-human CDH19 mAbs (Ca19Mabs) were developed through flow cytometry-based high-throughput screening. One clone, Ca19Mab-8 (IgG1, κ), specifically recognized CDH19-overexpressing Chinese hamster ovary-K1 cells but did not bind to other 21 CDHs (including both type I and type II) in flow cytometry. Additionally, Ca19Mab-8 recognized endogenous CDH19 in the human glioblastoma cell line LN229. The dissociation constant (KD) of Ca19Mab-8 for LN229/CDH19 was 9.0 × 10⁻⁹ M. Ca19Mab-8 can detect CDH19 in Western blotting and IHC. These findings suggest that Ca19Mab-8 is versatile for basic research and has potential applications in clinical diagnosis and tumor therapy.

Abstract: Background: Oncology medication costs are increasing internationally; patient attitudes towards these costs remain unclear. Methods: A three-part cross-sectional questionnaire was distributed to patients with breast cancer to determine their attitudes towards oncology medication costs and to ex-plore potential patient acceptable methods to reduce these costs. Results: 321 patients were eligible for inclusion and 180 fully completed the questionnaire (56.1% response rate). Overall, 67.8% (N = 122/180) of patients found the costs presented in the questionnaire to be unacceptable. 92.2% (N = 166/180), 87.8% (N = 158/180) and 68.9% (N = 124/180) of participants found the costs of pembrolizumab, palbociclib and trastuzumab respectively to be unacceptable. 72.8% (N = 131/180) of patients indicated that they would like to be better informed about the societal costs of their cancer treatment and 81.1% (N = 146/180) of patients believed that reducing the costs of cancer treatment to society is important. There was a statistically significant difference in patient desires to be better informed of societal drug costs between those with early-stage breast cancer and those with metastatic disease (75.8% vs 47.4%, χ2 = 6.923, p = 0.009). Conclusion: These findings indicate that many Irish patients with breast cancer find the societal costs of oncology medications to be unacceptable, and many patients have a de-sire to be better informed of these costs.

Abstract: 3-Amino-4-hydroxybenzoic acid (3,4-AHBA) is a non-proteinogenic aromatic compound that functions as a key biosynthetic precursor for diverse secondary metabolites with pharmaceutical and industrial value. Microbial production of 3,4-AHBA offers a sustain-able alternative to petroleum-based chemical synthesis; however, metabolic complexity and trade-offs between growth and product formation constrain rational strain design. Here, genome-scale metabolic (GSM) modeling and flux balance analysis (FBA) were in-tegrated with targeted genetic engineering to elucidate and enhance 3,4-AHBA production in Streptomyces thermoviolaceus. A genome-scale metabolic model was constructed and ex-panded by incorporating the nspH–nspI gene operon, which encodes the 3,4-AHBA bio-synthetic pathway. In silico FBA predicted substantial rewiring of central carbon metabo-lism, with carbon flux redirected from glycolysis and the tricarboxylic acid cycle toward aspartate-derived intermediates and 3,4-AHBA synthesis, accompanied by reduced bio-mass-associated flux. Guided by these predictions, an engineered strain (St::NspHI) was developed and experimentally evaluated. Consistent with model predictions, the engi-neered strain exhibited lower growth rates and glucose uptake than the wild type, reflect-ing a metabolic burden. Nevertheless, 3,4-AHBA production was achieved exclusively in the engineered strain. Comparison of simulated and experimental fluxes revealed overes-timation by FBA, likely due to secondary metabolism and incomplete genome annotation. Overall, GSM-guided design enables optimization of precursor production.

Abstract: Background: In recent years, virtual consultations have emerged as a crucial approach for continuity of chronic care provision, indicating a promising avenue for the future of smart healthcare systems. However, reversions to in-person care highlight persis-tent limitations, despite notable advantages of remote modalities. In parallel, recent developments in artificial intelligence (AI) indicate the potential to enhance remote chronic care, but user perceptions of such assistance and the corresponding human factors remain underexplored. Objective: This mixed-methods study aims to better understand the virtual consulta-tion experiences and attitudes toward AI assisted tools in remote care among patients with noncommunicable chronic conditions and their healthcare professionals (HCPs). It conducts an in-depth examination of the associated human-computer interaction and usability elements of virtual consultations and of potential AI assistance. Methods: Public and Patient Involvement was integrated to run pilots and refine documentations. Semi structured interviews with patients (n=10), focus groups with HCPs (n=15), and an online survey (n=83) were conducted. Qualitative data was ana-lysed through a reflexive thematic approach. The survey comprised the Telehealth Usability Questionnaire (TUQ) and bespoke items on user AI views, and the data was used to triangulate the qualitative findings. Nonparametric Kruskal–Wallis tests and ε² effect sizes compared TUQ and AI views scores between current and former virtual consultation user groups. Results: Seven themes emerged from the qualitative data, which were supported by the quantitative findings. The mean TUQ total score of 90.6 (SD=15.0) indicates high usa-bility and user satisfaction, and there were no significant group differences (p >0.05; ε² = 0.002–0.032). There was a clear preference for hybrid models, while a lack of em-pathy was identified during remote interactions. Users were cautiously open to AI as-sistance, contingent upon transparency, human oversight, and data integrity. Views on AI assistance did not differ significantly across groups (p >0 .05; ε² = 0.005–0.065). Conclusion: Virtual consultations for chronic conditions are widely usable and ac-ceptable, particularly through hybrid approaches. Addressing empathic engagement, holistic patient status, and transparent AI integration can enhance clinical quality and user experiences during remote interactions. This study has also identified evi-dence-based assistive AI features that can potentially enhance virtual consultations. These insights can inform the co-design of evidence based virtual care platforms, poli-cies and supportive AI tools to sustain remote chronic care delivery.

Abstract:

The use of zinc oxide nanoparticles (ZnONPs) in agriculture has increased due to their biostimulant potential; however, their effects on plant chemical communication and associated microbial communities are still poorly understood. This study presents a multi-perspective analysis contrasting the effects of ZnONPs with those of conventional ZnO (Bulk) on Capsicum annuum seedlings grown in a substrate with concentrations of 50 and 500 mg kg⁻¹. The results reveal that, at high doses, the bulk material (B500) generated a higher foliar accumulation of zinc (128.7 mg kg⁻¹) than ZnONPs (NP500, 119.7 mg kg⁻¹), a phenomenon attributed to the agglomeration of nanoparticles in the soil matrix, which limits their root absorption. At the physiological level, a critical divergence was observed: while bulk ZnO stimulated the activity of the enzyme superoxide dismutase (SOD), ZnONPs caused severe inhibition of the same (93% reduction), compromising the enzymatic antioxidant machinery and forcing the plant to rely on non-enzymatic mechanisms, such as an increase in total phenols. The volatilomic profile revealed a specific metabolic disturbance induced by ZnONPs in the green leaf volatiles (GLV) pathway. A significant accumulation of hexanal and suppression of hexanol and hexyl acetate were detected, suggesting that the nanomaterial inhibited alcohol dehydrogenase (ADH). In addition, ZnONPs suppressed the emission of methyl salicylate (MeSA)—a key messenger in acquired systemic resistance—whereas the Bulk treatment increased its abundance to 41.7%. Finally, metagenomic analysis indicated that zinc stress restructured the phyllosphere microbiota, promoting the proliferation of Actinobacteria and eliminating sensitive taxa such as Spirochaetes. Taken together, these findings demonstrate that ZnONPs act as multifactorial stressors that not only alter internal metabolism but also silence chemical communication and remodel plant ecology.

Abstract: The escalating complexity of urban emergencies, driven by rapid urbanization and climate change, highlights the critical need for advanced emergency response systems. Traditional methods, reliant on manual judgment and fragmented information, struggle to meet demands for rapid, precise, and efficient incident management. While large language models (LLMs) offer potential, general-purpose LLMs exhibit limitations in information timeliness, domain expertise, multi-modal data integration, and decision support accuracy within smart city emergency response. To address these challenges, we propose GuardianMind, a novel multi-modal enhanced LLM system specifically engineered for smart city emergency response. GuardianMind integrates a powerful base LLM with specialized modules: a City Emergency Knowledge Retrieval component, a Smart City Knowledge Graph, a Real-time Data and Tools module, and a Public Information Search module. This architecture enables GuardianMind to effectively process and synthesize diverse, heterogeneous data streams, providing a holistic understanding of emergencies and generating professional, accurate, and actionable response suggestions. Through comprehensive experiments on a custom-built dataset, GuardianMind consistently outperforms state-of-the-art general LLMs, including leading commercial and open-source models, across critical dimensions of accuracy, professional depth, and timeliness, while maintaining excellent language fluency. An ablation study further validates the indispensable contribution of each integrated module. Our qualitative analysis demonstrates GuardianMind's capacity to deliver highly precise, context-rich, and immediately actionable intelligence, marking a significant advancement in intelligent urban crisis management.

Abstract:

Purpose: This paper aims to systematically synthesize academic research published between 2020 and 2025 that investigates environmental, social, and governance (ESG) ratings and scores, with a focus on their methodologies, comparative performance, and impact on firm outcomes. Design/methodology/approach: A systematic literature review (SLR) was conducted using the Lens.org scholarly database. A structured title search retrieved 334 open access journal articles published between 2020 and May 2025 containing the terms "ESG Score", "ESG Rating", or "ESG Rater". The PRISMA 2020 protocol guided the selection and screening process. Findings: The literature exhibits growing concern about the divergence among ESG ratings, the methodological opacity of rating providers, and the variable financial implications of ESG scores. Common themes include score disagreements, rating agency biases, and emerging models for standardizing ESG assessments. Originality: This review provides the most up-to-date synthesis of ESG rating literature, focusing exclusively on articles explicitly addressing ESG ratings or scores in their titles. It contributes clarity to the fragmented ESG measurement space by organizing findings around key methodological and evaluative debates.

Abstract: This paper investigates percolation transitions in a disordered L-C system composed of inductors and capacitors (non-dissipative reactive elements). These transitions occur between different percolating states, resulting in distinct, constant values of the effective conductivity. We employ an exact approach based on the rotational symmetry of two-dimensional DC equations. A new type of phase transition is identified for these non-dissipative systems by analogy to a topological transition. The characteristics of these transitions, which are analogs of topological invariants, are calculated. We propose that these transitions may be considered a classical analog to quantum transitions, such as the quantum Hall effect.

Abstract: Environmental, Social, and Governance (ESG) frameworks are increasingly reshaping urban planning, real estate, and territorial governance. Originally conceived as corporate disclosure tools, ESG criteria are now influencing land use, regeneration strategies, and policy frameworks across Europe and beyond. This systematic review synthesizes 197 articles published between 2020 and 2025 to examine how ESG adoption translates into spatial, institutional, and governance outcomes. The findings show that ESG functions simultaneously as a financial instrument, a planning paradigm, and a governance mechanism. While it enables capital mobilization, climate resilience, and participatory innovation, it also risks reproducing socio- spatial inequities such as green gentrification, peripheral exclusion, and uneven infrastructure investment. Case studies from Florence, Cyprus, Russia, and broader European contexts demonstrate both methodological advances—such as spatiotemporal clustering, GIS-based analysis, and digital monitoring—and persistent gaps in regulatory frameworks, score reliability, and territorial integration. The paper contributes to planning scholarship by proposing an integrated framework that links ESG adoption to spatial justice, sustainable infrastructure, and multi-level governance. Policy implications emphasize the need to broaden ESG assessment to territorial indicators, embed safeguards against displacement, and align financial instruments with measurable social outcomes. Future research should advance geographic diversification, methodological innovation, and normative engagement with equity and resilience.

Abstract: Road freight transportation remains the dominant mode for goods distribution worldwide, with articulated vehicles playing a critical role in this sector. However, these vehicles are prone to severe instability phenomena such as jackknifing, trailer sway, and rollover, particularly under high-speed or emergency maneuvers. This paper presents an advanced steering stability control strategy for articulated vehicles based on Model Predictive Control (MPC) and differential braking, aiming to enhance lateral and yaw stability during autonomous driving operations. The proposed controller integrates trajectory tracking and yaw stability objectives within a unified optimization framework, systematically handling multi-variable constraints. A dynamic model of a tractor–semitrailer combination has been developed, enabling accurate representation of vehicle kinematics and tire forces. Simulation results demonstrate that the inclusion of differential braking significantly reduces articulation angle and yaw rate deviations, preventing instability even at speeds exceeding the critical threshold of 31.04 m/s. Comparative analysis reveals that coordinated braking applied to both tractor and trailer units achieves superior performance over single-unit application, particularly under high-speed conditions. While the findings confirm the effectiveness of MPC-based differential braking for articulated vehicle stability, the study also highlights the current limitation of simulation-based validation and the need for experimental testing to ensure real-world applicability. Future research should explore multi-actuator coordination, including active front steering integration, to further enhance stability and reduce longitudinal speed loss.