Volume 8.
Issue 03.

The article is devoted to the study of the proportion of contamination of raw cotton, the geometric dimensions and the number of contaminations of raw cotton are analyzed. It is shown that the existing method used in practice for determining the contamination of raw cotton by the weight of the litter does not fully characterize the contamination of raw cotton as an object of cleaning. The influence of the type and quality of the collection and the variety of raw cotton on the weed fraction was determined.

It is shown that in the composition of raw cotton of the highest grades and machine, about the manual collection, there is a significant amount of collection of small litter.

It has been established that weed impurities up to 6 mm in size make up 93-97% of the total contamination of raw cotton.

It is recommended, based on the analysis of the results of the experiment, to improve the quality of the collection of raw cotton, control the collection, take into account the influence of fractional cotton harvesters, take into account the fractional composition of weeds when drawing up a plan for harvesting raw cotton, and also carry out a wide development of the composition of raw cotton in this area by studying the factors influencing the study of the geometric dimensions of contamination, give the appropriate conclusions.

Modern artificial intelligence and high-performance computing systems face critical bottlenecks in inter-chip communication as computational capabilities continue to advance beyond the limits of traditional copper-based interconnects and board-edge optical modules. Co-packaged optics emerges as a transformative solution by integrating photonic components directly within processor packages, dramatically reducing electrical path lengths and enabling unprecedented bandwidth densities while lowering energy consumption per transmitted bit. This article presents a comprehensive hybrid integration architecture that combines two-and-a-half-dimensional and three-dimensional packaging techniques to co-locate photonic chiplets with compute dies on silicon interposers. The article leverages micro-ring resonator-based wavelength division multiplexing on silicon photonic platforms to achieve high aggregate throughput while maintaining compatibility with advanced logic manufacturing processes. Through detailed co-design of packaging structures, electrical-optical interfaces, thermal management systems, and control algorithms, the architecture addresses key technical challenges that have historically impeded photonic integration efforts. Validation through multi-level simulations and physical prototypes demonstrates feasibility for rack-scale optical connectivity meeting the demanding requirements of distributed machine learning workloads. The article examines critical parameters that affect the timing of commercial adoption, such as manufacturing yield, serviceability, the need for standardization, and economic trade-offs. Results indicate that hybrid co-packaged optics architectures provide viable pathways for sustaining bandwidth scaling in next-generation data center fabrics serving artificial intelligence applications.

Zenodo DOI:- https://doi.org/10.5281/zenodo.18875724

Physically unclonable functions (PUFs) convert microscopic manufacturing noise inside a chip into a device-specific hardware fingerprint, but delay-based designs fall to machine-learning modelling once an adversary collects enough challenge-response pairs (CRPs). This article reports an empirical study of six machine-learning attacks — logistic regression, least-squares regression, a multilayer perceptron, an LMN feature-lifting model, a denoising autoencoder, and a compact Transformer encoder — against Arbiter OR-AND-XOR PUFs (AOX-PUFs), simulated with pypuf on the University of Sheffield's Stanage high-performance computing cluster across CRP budgets from 2,500 to 1.9 million. The attacks exploit the Arbiter PUF's linear-additive delay model, and the defence under test is a BB84-inspired encoding wrapper motivated by the Quantum Lock forging-probability bound. Classical AOX-PUFs were modelled with 95–99% accuracy within tens of thousands of CRPs, while BB84-inspired encoding capped every attacker at a flat ceiling of roughly 0.85–0.90 that did not rise within either batch, even at close to two million CRPs. Decomposing the encoding pipeline shows that this ceiling tracks the simulated measurement-success probability rather than a genuine basis-secrecy effect: disclosing the encoding basis altered outcomes by less than a percentage point. An illustrative calculation based on the Quantum Lock bound shows that a genuine quantum advantage should decay the forging probability exponentially as the number of encoded output qubits grows, a prediction the flat empirical ceiling does not fulfil. The results corroborate prior architectural-hardening findings for AOX-PUFs while diverging from the CRP-inflation behaviour the Quantum Lock model predicts.

This article examines the application of artificial intelligence (AI) models and methods in the selection and evaluation of innovative projects. With the increasing complexity of decision-making processes in innovation management, traditional evaluation techniques often fail to ensure objectivity, scalability, and predictive accuracy. The study analyzes machine learning algorithms, decision-support systems, and multi-criteria evaluation models used in project selection. The methodological framework is based on a systematic review of academic literature and empirical findings from recent studies. The results demonstrate that AI-based approaches significantly improve decision quality, reduce uncertainty, and enhance predictive capabilities. The article also discusses the limitations and challenges associated with AI implementation in innovation evaluation processes.

The high growth rate in digital businesses has heightened issues around the world about the environmental footprint of information technology (IT) infrastructure, especially as the data centres, cloud services, and high-performance computing workloads are becoming a large contributor to increasing energy use and carbon emissions. Although the topic of sustainability has gained increased regulatory and corporate attention, there remains no set, data-based approaches to measure the ecological performance of the organizations with regard to the digital operations. The paper presents a holistic analytical model that combines sustainable IT infrastructure indicators, cloud resource optimization policy, and corporate sustainability key performance indicators (KPIs). Based on a mixed-methodology, which integrates empirical data related to industry benchmarking, cloud provider sustainability reporting, and already existing environmental reporting criteria, the framework provides a systematic approach to connect micro-level IT energy telemetry (including power usage effectiveness (PUE), server utilization rates, virtualization efficiency, and carbon intensity of workloads) to macro-level sustainability results, including reductions in greenhouse gas (GHG) emissions, integration of renewable energy, energy cost reductions, and improvement in ESG performance. Quantitative modeling based on real world data demonstrates how green data analytics can be used to aid in carbon-conscious scheduling, predict IT energy demand and optimizing allocation of cloud resources. The outcomes indicate that the incorporation of IT operating data with high-end analytics can improve greatly the level of transparency, the measurement precision, and the environmental responsibility of digital enterprises. The originality of the study is the cross-layered mapping of technical IT metrics on organizational sustainability KPIs, which can be replicated to achieve the net-zero digital transformation. The suggested framework offers practical information to businesses, policymakers, and cloud service providers interested in realizing sustainability goals in fast-changing digital realms.

This article presents a systematic analysis of the geological structure and hydrogeological characteristics of the Isfara underground water deposit, which constitutes an integral part of the Fergana hydrogeological basin. The study provides a detailed examination of the Quaternary period sediments of the region, specifically the stratigraphy and lithological composition of the Sokh, Tashkent, Mirzachol, and Syrdarya complexes. The results of long-term scientific investigations concerning the hydrodynamic regime of the Isfara River alluvial fan, the recharge sources of groundwater, and the discharge balance are comprehensively summarized.

In the current work, the researchers examine 20 years (2004-2024) of literature on Industry 4.0 and IoT, visualizing intellectual organisations, knowledge sharing, and changes in themes. Based on data collected from 6,000+ articles scraped from Scopus and processed with VOSviewer, Biblioshiny, and Python, it answers three research questions: the creation of Industry 4.0 and IoT scholarship, market forces and market uptake, and convergences in technologies. Results indicate soaring growth in publications starting in 2016, with significant contributions from China, the US, Germany, India, and the UK. The study identifies a gap in the scholarly literature and in the uptake of IoT, with high implementation evident in the manufacturing and logistics sectors, whereas in the healthcare, agriculture, and building sectors, implementation challenges persist. It includes suggestions for what researchers should consider in the future, including socio-technical integration, cybersecurity, and sustainable digital ecosystems.

A recent paper from Thinking Machines Lab (TML), "Defeating Nondeterminism in LLM Inference," has provided a new perspective on the prevalence of nondeterministic outputs in Large Language Models (LLMs) configured for deterministic behavior.[1] This issue undermines reliability, complicates testing, and hinders scientific reproducibility, with studies showing accuracy variations of up to 15% across identical runs.[2] This paper's primary contribution is to analyze the TML findings through a novel three-part framework, categorizing the boundaries of any determinism solution as: (1) an operational boundary (reproducibility is local to a specific hardware/software stack); (2) a functional boundary (it applies only to greedy decoding, not generative sampling); and (3) an architectural boundary (it does not solve nondeterminism in distributed, multi-GPU systems). This analysis argues that the TML work provides a critical engineering trade-off for reproducibility rather than a complete solution to nondeterminism. By situating the TML work within the proposed framework, this analysis clarifies what is practically achievable versus what is fundamentally impossible in the pursuit of deterministic AI.

The current financial regulatory frameworks, specifically the ‘Fundamental Review of the Trading Book’ (FRTB), require almost real-time disclosure of Expected Shortfall and Value at Risk measures of heterogeneous portfolios of equities, fixed-income securities, and complex over-the-counter derivatives. Switching from batch-based reporting to continuous intraday risk reporting poses significant computational challenges, especially when dealing with portfolios of over 100,000 positions with nonlinear sensitivities. This article introduces the Decentralized Risk Computation Grid (DRCG), a cloud-native design that uses stateful orchestration and sensitivity-aware sharding to decentralize portfolio decomposition to elastic worker clusters. In contrast to the classical stateless parallel models, the DRCG uses a warm-cache worker pattern that avoids unnecessary market data loading cycles, and tail latency is significantly lower, and yet deterministic recovery is achieved in the event of node failures. The framework is horizontally scalable, with sensitivity-based portfolio decomposition formalized mathematically and Byzantine fault-tolerance principles, without affecting audit compliance or data consistency. Empirical confirmation in distributed settings shows that state-conscious orchestration offers the resilience required in systemic risk surveillance in times of high market volatility and stress.

Zenodo DOI:- https://doi.org/10.5281/zenodo.18875781

The evolution of digital protection mechanisms between 2022 and 2025 reflects a critical convergence of cybersecurity, power system protection, and policy-driven governance frameworks. This study presents a comprehensive multi-year assessment of digital protection trends, focusing on four primary dimensions: human capital, policy structures, threat exposure, and capability evolution. Drawing exclusively from established literature in power system protection, simulation-based testing, and policy-based system management, alongside longitudinal cybersecurity observations, this research synthesizes theoretical and applied perspectives into a unified analytical framework.

The study identifies a paradigm shift from hardware-centric protection mechanisms to software-defined, simulation-driven, and policy-controlled protection architectures. Historical contributions such as electromagnetic transient simulations (Dommel, 1969) and real-time relay testing frameworks (Kezunovic et al., 1994) are examined as foundational elements that have influenced contemporary digital protection strategies. Policy-based management systems and privacy-aware architectures further demonstrate the growing complexity of governance in distributed and autonomous environments (Beigi et al., 2004; Karat et al., 2005).

A key contribution of this research lies in its integration of cybersecurity workforce trends and governance maturity insights, particularly those highlighted in recent longitudinal studies (Thanvi, 2026). These insights reveal persistent skill gaps, evolving threat landscapes, and the increasing importance of adaptive capability frameworks. The analysis further explores the role of simulation technologies and real-time systems in enhancing resilience against emerging threats.

Findings indicate that while technological capabilities have advanced significantly, organizational and policy-level adaptations lag behind, creating systemic vulnerabilities. The study proposes a conceptual model linking human expertise, policy enforcement, and technological capability as interdependent pillars of effective digital protection.

This research contributes to academic and practical discourse by offering a multi-dimensional evaluation framework and identifying critical gaps in current protection paradigms. It concludes with recommendations for integrating simulation-based validation, policy automation, and workforce development to achieve robust and adaptive digital protection systems.

Digital Twin technology has become one of the key component of Industry 4.0 and smart manufacturing systems. The concept enables real-time synchronization between physical industrial assets and their virtual representations. This research investigates the architecture, mathematical models, and predictive algorithms used in Digital Twin systems for industrial automation. A comprehensive mathematical framework based on dynamic system modeling, data-driven analytics, and predictive maintenance algorithms is proposed. Experimental analysis demonstrates that the implementation of Digital Twin technology improves monitoring accuracy, reduces downtime, and increases production efficiency. The proposed system integrates sensor networks, industrial IoT, cloud computing, and machine learning algorithms to optimize industrial operations.

The article examines the architectural role of the Model Context Protocol in the design of multi-agent systems intended to support robust coordination between cognitive and infrastructural components. The relevance of the study is determined by the rapid transition from isolated LLM interactions to distributed agentic environments, in which direct coupling between agents and external APIs, data sources, and tools increases system interdependence, complicates maintenance, and weakens governability. The purpose of the article is to provide a conceptual substantiation of the architectural principles of multi-agent systems based on MCP and to identify their significance for scalability, interoperability, security, and maintainability. The scientific novelty of the study lies in interpreting MCP not merely as an auxiliary integration protocol, but as an autonomous architectural layer that establishes a formalized topology of interaction among hosts, clients, and servers. The principal findings demonstrate that the use of MCP ensures a strict separation of concerns, declarative context management, modular composability, and governance-by-design in capability access. The article concludes that such an approach enhances observability, reduces integration fragility, and creates the foundation for the evolutionary development of multi-agent solutions while preserving control and auditability. The article will be useful for researchers, AI system architects, developers of agentic platforms, and specialists in enterprise digital infrastructure.

The article is dedicated to the study of methods for automated detection and localization of defects in program code within continuous integration environments. The relevance of the research is determined by the growing complexity of distributed software architectures and the increasing dependence of development processes on continuous testing pipelines. The novelty of the work lies in the integrated analytical examination of detection stability, regression test selection, prioritization strategies, spectrum-based localization, trace reconstruction, and slicing refinement within a unified CI diagnostic loop. The work describes mechanisms for writing automated tests that verify login flows, payment operations, interface elements, forms, and APIs, as well as approaches to executing these tests on every code change to ensure early defect exposure and regression protection. Special attention is paid to the interaction between algorithmic ranking models and developer feedback in practical debugging scenarios. The study sets itself the goal of systematizing engineering mechanisms that enhance diagnostic precision under time constraints. Comparative analysis and source synthesis methods are used. The conclusion substantiates that adaptive coordination of testing layers increases reliability. The article will be useful for software engineers, QA automation specialists, and CI architects.

In this work, the physicochemical regularities of the heterogeneous decomposition process of dolomite from the Navbakhor deposit in a nitric acid environment were studied. In the study, complexometric titration and X-ray diffractometry methods were used, and the dependence of the reaction medium on concentration and temperature factors was analyzed. It was established that in the optimal process mode (120% stoichiometry, 35% HNO3), active dissolution of carbonate minerals is observed, and in the solid phase, mainly inert quartz and wollastonite phases are concentrated. X-ray diffraction analysis of the solid residue formed after acid decomposition showed that the main part of the carbonate minerals was dissolved during the reaction, and it was found that mainly quartz (73.6%), wollastonite (22.4%), and magnesium-aluminosilicate phases (4.0%) remained in the solid phase.

The obtained results have scientific and practical significance for improving the technology for obtaining magnesium-containing solutions and producing magnesium compounds based on the acid treatment of local dolomite raw materials.

Digital twin technology has emerged as one of the most transformative paradigms in modern cyber-physical systems, enabling the creation of dynamic virtual representations of physical assets, environments, and processes. When integrated with next-generation communication infrastructures and edge intelligence, digital twins enable real-time monitoring, predictive analytics, and autonomous decision-making across domains such as manufacturing, aerospace, smart cities, and unmanned aerial systems. The convergence of digital twin architectures with edge computing and artificial intelligence is particularly relevant in environments requiring ultra-low latency and high-fidelity simulation, including autonomous vehicles, industrial automation, and distributed drone systems. However, the deployment of real-time digital twin platforms introduces significant challenges related to interoperability, security, privacy, and cross-domain standardization. Emerging communication networks, including 5G and anticipated 6G systems, promise to address these challenges by enabling scalable, high-bandwidth connectivity and distributed intelligence.

This study investigates the evolving role of secure edge intelligence in enabling scalable digital twin deployments within next-generation communication ecosystems. Drawing from a comprehensive analysis of prior research across smart manufacturing, autonomous aerial systems, industrial automation, and edge computing architectures, the research synthesizes theoretical frameworks that explain how distributed intelligence can support real-time synchronization between physical systems and their digital counterparts. Particular attention is devoted to digital twin implementations in autonomous vehicles, unmanned aerial vehicles, smart grids, and industrial production environments. The analysis explores the integration of machine learning models, distributed edge computing platforms, and next-generation wireless technologies to support digital twin operations at scale.

The study further examines security and privacy implications associated with edge-enabled digital twin systems, highlighting vulnerabilities arising from distributed data flows, device heterogeneity, and cyber-physical integration. Through an extensive theoretical analysis of contemporary research, the article proposes a conceptual framework that integrates edge intelligence, cross-domain standardization, and secure communication protocols for digital twin environments. The findings suggest that the future of digital twin ecosystems will rely heavily on intelligent edge architectures capable of supporting autonomous decision-making while ensuring trustworthiness, resilience, and interoperability across complex technological infrastructures.

Current microphysiological systems lack predic- tive fidelity for human pharmacokinetics due to oversimplified mathematical representations that conflate biological processes. We introduce a physics-informed digital twin framework that deconvolves active metabolic clearance from passive drug trans- port phenomena in liver-on-chip platforms. Our computational architecture employs a three-compartment ordinary differential equation model mapping media, interstitial, and intracellular domains, integrating hardware-specific microfluidic constraints with compound physicochemical properties. The framework was validated across 32 compounds spanning multiple hepatic microphysiological systems, demonstrating superior predictive accuracy with a mean clearance ratio of 1.04 ± 0.31 versus 0.56 ± 0.44 for conventional single-compartment models. By disentangling permeability, partitioning, and metabolic pathways, the digital twin enables mechanistic interpretation of observed kinetics while maintaining compatibility with standard experi- mental protocols. The open-source R implementation facilitates seamless integration with physiologically-based pharmacokinetic modeling for clinical translation. This paper establishes a compu- tational foundation for precision drug development using organ- on-chip technologies. 

Healthcare integration engines such as Mirth Connect (NextGen Connect) are central to clinical interoperability, enabling the exchange of HL7, FHIR, and proprietary messages between disparate healthcare systems. A critical operational hazard arises when these engines invoke external third-party APIs that become slow, unresponsive, or permanently unavailable. Without explicit timeout controls, integration channels may block indefinitely, causing thread exhaustion, message backlog accumulation, degraded throughput, and complete engine unavailability—outcomes that directly threaten patient care continuity and regulatory compliance. This paper proposes a fault-tolerant timeout framework for external service calls in healthcare integration engines. The framework combines configurable connection and read timeouts, selective retry with exponential backoff, circuit breaker state management, and graceful fallback handling into a centralized invocation layer. Each component is governed by principles of deterministic termination, failure isolation, configurability, auditability, and clinical safety. Pseudocode algorithms are provided for each mechanism. Although implemented in the context of Mirth Connect, the architecture generalizes to other synchronous and asynchronous integration platforms. Evaluation demonstrates that the framework significantly reduces the risk of indefinite channel blocking while preserving message traceability, data integrity, and compliance with healthcare data governance requirements.