Adaptive and multi-functional nozzle actuation is highly demanded for modern aero-propulsion systems. Realizing integrated thrust vectoring, variable exit-area adjustment and reverse-thrust generation within a compact layout still poses substantial challenges. Inspired by archerfish’s integrated flow-regulation mechanism, which relies on coordinated oral bones and muscles to generate asymmetric oral deformation for jet shaping, direction control and flow-passage switching, this paper develops a nozzle actuation mechanism drawing on the kinematics of the archerfish jaw-operculum system. The design integrates pitch vector adjustment, variable exit-area modulation (full-closure included), and reverse-thrust generation. A single electric-cylinder-linkage assembly drives the nozzle exit, while lead-screw actuators govern the operculum-mimicking flow-diversion structure for reverse-thrust switching. Validated by kinematic analysis and three-dimensional flow-field simulations, the mechanism delivers a maximum continuous geometric pitch deflection of 36° within 8 s and stepless exit-area adjustment ranging from 1.8 × 104 mm2 to full closure. It enables active modulation of jet mixing behaviours and reverse-thrust output for landing deceleration. Compared with conventional nozzles, this bionic configuration achieves a streamlined actuation layout, competitive vectoring performance, continuous area adjustability and reduced control complexity, offering an innovative bionic solution for adaptive aero-propulsion devices.
Cross-platform cultural heritage systems must coordinate structured domain data, shared services, real-time 3D representation, and device-specific interaction. This paper presents a metadata-centered architecture for delivering a numismatic collection through responsive web, map, conventional 3D, and immersive WebXR interfaces. It combines a relational metadata and media layer, shared REST-based services, and device-adaptive clients. It was instantiated with 253 historical coins. Instead of storing a coin-specific 3D model for each record, Babylon.js generates lightweight geometry at runtime and maps paired obverse and reverse images onto its surfaces. Record identifiers, metadata, media references, and query logic are reused across desktop, mobile, and head-mounted display access. An implementation-level check confirmed that an updated record was available through the catalog, map-related view, and 3D/WebXR gallery without client-specific duplication. Complementary formative evaluations involved 97 web participants and 19 WebXR participants, with mean System Usability Scale scores of 80.41 and 83.55, respectively. Feedback identified object manipulation, information legibility, visual fidelity, and interaction comfort as refinement priorities. This study contributes a system-level approach that connects shared cultural heritage services with conventional and immersive clients without separate collection databases or content-management workflows.
Reliable wearable motion tracking depends not only on nominal sensor precision but on whether calibration remains valid after re-donning, anatomical misalignment, magnetic disturbance, temperature change, and prolonged operation. This structured critical review treats calibration as a system-level engineering decision spanning intrinsic inertial measurement unit (IMU) correction, sensor-to-segment alignment, flexible-sensor calibration, drift management, and complementary or external references. A reproducible Scopus core search conducted in May 2026 identified 807 records. Three co-authors screened titles, abstracts, and keywords using explicit eligibility criteria, with full-text inspection when required; the final 110-source reference corpus combines eligible core records with supplementary and contextual sources identified through coverage checks in IEEE Xplore, Web of Science, ScienceDirect, and the MDPI database. Unlike prior reviews that mainly treat individual calibration layers or inertial-error mechanisms, this review jointly compares sensor-level, anatomical, multimodal, adaptive, and external-reference routes within a common deployment-oriented framework. Evidence is coded by failure mechanism, calibration route, sensing modality, application context, validation context, and deployment constraint. No universally superior route is supported: static and functional alignment differ in burden and anatomical sensitivity; magnetometer use depends on field reliability; vision and ultra-wideband (UWB) references can restore absolute observability but add infrastructure dependence; and direct IMU–flex calibration remains comparatively sparse. The resulting decision framework prioritizes reproducibility, robustness, calibration stability, and fitness for real-world deployment over isolated best-case accuracy.
Reliable temperature estimation in GaN high-electron-mobility transistors requires temperature-sensitive electrical parameters (TSEPs) that combine sufficient sensitivity with stable calibration behavior. This study compares on-state resistance (Rds(on)), threshold voltage (Vth), reverse-conduction voltage (Vsd), and gate-source voltage (Vgs) using three nominally identical 650 V enhancement-mode GaN HEMTs at three reference temperatures of 25, 50, and 75 °C. Each TSEP was repeatedly measured at each reference temperature, and the median value of the repeated measurements was used for calibration. Calibration performance was primarily assessed using temperature sensitivity, linearity, short-term repeatability, cross-device consistency, and leave-one-device-out calibration transferability, while a limited leave-one-temperature-out analysis was retained only as an additional internal-consistency check. Rds(on) showed a mean K-factor magnitude of 0.247 mΩ/°C, a mean R2 of 0.9974, and a K-factor coefficient of variation of 1.60%. Vgs provided the highest voltage-domain sensitivity, with a mean K-factor magnitude of 2.053 mV/°C, a mean R2 of 0.9980, and a coefficient of variation of 1.49%, whereas Vth and Vsd exhibited K-factor coefficients of variation of 5.93% and 46.11%, respectively, with the latter showing pronounced cross-device variation. The leave-one-device-out analysis yielded transfer RMSEs of 5.600 °C for Rds(on) and 6.324 °C for Vgs, substantially lower than 55.368 °C for Vth and 94.631 °C for Vsd, demonstrating that cross-device slope consistency and calibration transferability are related but distinct characteristics. These results demonstrate that robust TSEP selection should consider repeatability, cross-device consistency, and calibration transferability in addition to sensitivity and linearity.
Extracting Cole–Cole relaxation parameters from broadband dielectric spectroscopy (BDS) of cross-linked polyethylene (XLPE) cable insulation is ill-conditioned when the dispersion peak lies outside the instrument frequency window. This paper proposes an identifiability-aware, physics-constrained machine-learning framework and quantifies the analytic bound on what is recoverable over [0.01, 105] Hz. A Cramér–Rao lower bound (CRLB) analysis indicates that under the assumed model and noise conditions, ε∞ and σdc are well identified (relative CRLB ≤ 0.05%), whereas εs and log τ remain weakly identifiable (|ρ| = 0.97). α is strongly correlated with log τ but retains a comparatively small marginal CRLB. A physics-based sim-to-real strategy generates 10,000 synthetic spectra; the domain gap against four reconstructed real spectra is measurable but not conclusively resolved at this sample size. A label-free physics-constrained spectral optimisation (PCO) step reduces the relative εs deviation at 140 °C from 37.8% to 8.8%, an improvement specific to the analysed spectrum that Monte Carlo analysis does not show to generalise across noise realisations; PCO does not improve on a 50-start Levenberg–Marquardt baseline at 180 °C, where residual deviations indicate estimator bias and/or model–data mismatch rather than an identifiability limit. A measurement design criterion shows that a 5% CRLB-based uncertainty target for log τ requires fmin ≤ 10−3 Hz. Because the assessment rests on a single digitally reconstructed dataset, the results constitute a methodological proof of concept rather than broad validation.
To address the challenge of limited evaluations of expensive high-complexity functions in surrogate modeling, this paper proposes a methodological framework that integrates multi-fidelity Gaussian process (MFGP) regression, sequential active sampling, and global sensitivity analysis, using a radar angle tracking accuracy (ATA)-inspired synthetic function as the validation benchmark. An autoregressive MFGP model with composite kernel functions is constructed to capture the complex nonlinear characteristics of the synthetic response surface. A sequential active sampling strategy based on the maximum uncertainty criterion, combined with fixed resource allocation and a neighboring sample avoidance mechanism, is designed to achieve efficient sample collection under a limited evaluation budget. Sobol global sensitivity analysis is then performed on the trained MFGP surrogate model to quantify the main and interaction effects of each input parameter on the synthetic ATA function. Experimental results demonstrate that with the complete sequential sampling procedure, the proposed MFGP model achieves an R2 of 0.8515 and an RMSE of 0.0850 on the test set, significantly outperforming the single-fidelity GP model that relies solely on high-complexity samples. Sobol analysis identifies the Jamming-to-Signal Ratio (JSR) and lateral distance as the most critical influencing factors within the synthetic benchmark. The proposed framework substantially improves sample efficiency under limited evaluation budgets, providing an effective methodological reference for surrogate modeling in similar high-cost computational scenarios.
This article introduces an experimental approach to evaluating the thermal resistance of industrial resistance temperature sensors and demonstrates their use for estimating the flow rate of the surrounding medium. A new method, the Self-Heating Effect Anemometry (SHEA) method, is proposed. Experimental studies were conducted for various flow rates (up to 6 m/s) and excitation current values (up to 5 mA) to evaluate the thermal properties of the sensors under dynamic operating conditions. The results obtained show that the proposed method can be effectively used to select an appropriate measurement current that minimises the self-heating effect and keeps the temperature measurement error within an acceptable range. At the same time, the method allows one to determine the flow rate of the surrounding medium from the measured thermal resistance of the sensor. The presented approach can be applied in both precise temperature measurements and flow monitoring systems using thermoresistive sensors. The proposed method simultaneously provides measurements of both the temperature and the flow rate of the surrounding medium. The research contributes to a better understanding of the self-heating effect in industrial measurement applications and provides practical guidance to improve measurement accuracy and sensor operating conditions.
Lane Keeping Assistance (LKA) systems play a critical role in enhancing vehicular safety and driving comfort by maintaining lane alignment and mitigating risks associated with driver distraction or drowsiness. These systems rely on sensor data to execute corrective steering or braking actions, yet their dependence on interconnected electronic components exposes them to a range of safety and cybersecurity threats. Attackers can exploit vulnerabilities in sensors, communication protocols, and Electronic Control Units (ECUs), potentially triggering false interventions or disabling safety functions. This paper presents a comparative Threat Analysis and Risk Assessment (TARA) of two LKA system architectures using the Medini Analyze tool. The first architecture employs a hierarchical controller with driver-intention detection and Electronic Stability Control (ESC)-based actuation. The second architecture employs a Learning-Based Model Predictive Control (LBMPC) framework enabling situation-adaptive decision-making. Through systematic identification and evaluation of threats and vulnerabilities, the analysis assesses risk levels associated with each design. The comparative analysis reveals trade-offs among architectural complexity, system robustness, and exposure to potential vulnerabilities, offering practical insights to improve the safety and security of LKA system designs.
To address rapid frequency decline, transient voltage violations, and excessive configuration costs caused by competition between active and reactive power support in weak grids, this paper proposes an optimal configuration method that incorporates the fault-period voltage-support capability of photovoltaic (PV) inverters into the planning of a grid-forming energy storage system (ESS). A coordinated response model for the ESS and PV inverter is developed, in which the PV inverter provides reactive power through Q-V droop control while smoothing its active power output. An optimization model is then formulated to minimize the annualized ESS cost while satisfying constraints on transient frequency security, voltage recovery, islanded operation, and state of charge (SOC). Frequency security indices, including the rate of change in frequency, frequency nadir, and quasi-steady-state frequency deviation, are explicitly linked to the rated power and energy capacity of the ESS. A hierarchical solution framework integrating capacity search, scheduling while connected to the grid, stepwise transient verification, and steady-state assessment under islanded operation is adopted to improve computational efficiency. Case studies on a weak distribution network show that PV transient voltage support reduces the reactive power requirement of the grid-forming ESS and lowers its configuration cost by approximately 5.2%. Meanwhile, all frequency and voltage indices remain within the prescribed security limits. Further multi-scenario evaluations and sensitivity analyses confirm the broader applicability of the proposed method across different operating conditions.
This paper presents a fast electrothermal coupled analysis method that combines a temperature-dependent loss model with an analytical multilayer thermal model for multiple heat sources to predict the junction temperatures of silicon carbide metal-oxide-semiconductor field-effect transistors (MOSFETs) cooled by a heat-pipe heat sink. Mori–Tanaka homogenization is used to represent the heat-pipe heat sink as an equivalent anisotropic medium, and the temperature field is obtained using a Fourier-series analytical solution. Thermal coupling among four SiC MOSFETs is captured by an influence-coefficient matrix to update junction temperatures via a simple matrix–vector product during electrothermal iterations. The proposed method is validated through PSpice and ANSYS Icepak simulations. The maximum junction-temperature error is 2.17 °C, and the total computation time is reduced from 8134 to 75 s, corresponding to an ~108× speedup. These results confirm that the proposed method is an accurate and computationally efficient junction-temperature prediction tool for the iterative thermal design of high-power-density converters.
As DevOps pipelines accelerate release cycles from weeks to hours, testing practices face a fundamental misalignment: most AI-driven approaches automate isolated tasks rather than orchestrating testing activities across the delivery lifecycle. Although generative AI has broadened the capabilities of software testing, this task-centric focus leaves a critical autonomy gap unaddressed. To characterize this gap, this study presents a systematic literature review (SLR) of AI-driven software testing in DevOps environments. Through a structured and reproducible methodology that assessed study quality, 31 articles were selected from Scopus and Web of Science. The SLR evaluates the level of autonomy achieved by existing approaches and examines their integration within continuous integration and continuous delivery (CI/CD) pipelines. Within the journal-based corpus captured by the predefined review protocol, 27 of the 31 articles (87%) operate at task-level assistance, 3 (10%) demonstrate stage-level integration, only 1 (3%) exhibits partially agentic behavior within a bounded domain, and none demonstrates lifecycle-wide, goal-driven orchestration. Two additional shortcomings emerge from the literature: the absence of standardized criteria for evaluating testing autonomy and the limited empirical validation of proposed approaches in operational CI/CD environments. Based on these findings, this study introduces a four-level autonomy taxonomy for AI-driven testing and outlines the architectural, operational, and governance conditions required for future Agentic TestOps systems, understood as goal-driven and self-adaptive testing systems operating across delivery pipelines. Together, these contributions help define a research agenda for advancing DevOps testing toward more autonomous testing approaches.
Efficient operation of interconnected production systems requires accurate modeling of upstream–downstream interactions to support informed operational decision-making. This study proposes an interconnected Adaptive Neuro-Fuzzy Inference System (ANFIS) framework consisting of two sequentially linked ANFIS models representing the upstream manufacturing stage and the downstream packaging stage, where the output of the first model is incorporated as an input to the second model. The framework uses operational Key Performance Indicators (KPIs), including mean time between failures (MTBF), mean time to repair (MTTR), uptime, reject rate, short stops, and long stops, to capture the nonlinear relationships between manufacturing and packaging operations. Unlike conventional single-model approaches, the proposed methodology represents the manufacturing and packaging stages as interconnected neuro-fuzzy subsystems, explicitly modeling their operational dependencies while maintaining model interpretability. Following data preprocessing and conditioning, two ANFIS models were developed using real industrial data and integrated into a MATLAB/Simulink environment for scenario-based analysis. The developed ANFIS models achieved low testing errors and satisfactory predictive performance on real industrial data. Simulation-based analyses were subsequently conducted to evaluate the effects of stoppage behavior and maintenance-related improvements on production-line uptime. The results indicate that the proposed framework captures upstream–downstream production dependencies and provides an interpretable decision-support tool for evaluating operational improvement scenarios in manufacturing environments.
Electric cars seem to be the future of transportation, capturing a market share that can no longer be neglected. The complete transition to electromobility, like any technological shift, is not without challenges. The critical constraint remains the battery, which, despite recent significant advances, remains the primary limitation of the electric car. The paper proposes a comparison between battery electric vehicles (EVs), internal combustion engine vehicles (ICEVs) equipped with spark-ignition engines, and hybrid electric vehicles (HVs) under different operating conditions. Given the scope of the subject, the paper presents a comparative analysis of the performance, advantages, and disadvantages of these three types of vehicles.
In recent years, machine learning algorithms are increasingly dependent on large volumes of data for their training, including personal data, while at the same time the law has strengthened the right of individuals to have such data deleted, thus creating an inherent tension. Regulations such as the General Data Protection Regulation (GDPR) oblige an organization to erase personal data on request, but deleting a record from a database is not enough. A trained model retains the influence of that record in its parameters and may still expose it, for example, through membership inference. Machine unlearning has emerged in order to remove this influence from the model itself, and it has rapidly developed into an active research area. However, existing surveys have not provided a unified, verifiability-centered account of what is required to demonstrate that unlearning has actually occurred. This review provides a unified treatment of machine unlearning, beginning with the taxonomy of exact and approximate algorithms and the trade-off between efficacy, fidelity, and efficiency that governs them. It then examines the role of unlearning in privacy protection and its dual role in security, where it serves as a defense against poisoning and backdoors but also becomes an attack surface. Particular attention is given to evaluation, because the empirical tests of the literature can measure a removal but cannot prove it. On this basis, the review examines verifiable, federated, and decentralized unlearning, including the Proof of Unlearning and zero-knowledge constructions. Taken together, the review’s findings indicate that most methods assert rather than prove removal, while verifiable unlearning in federated and decentralized environments remains a central open problem.
Effective and efficient supply chain management is important for the healthcare sector to access essential medications and medical products on time. Blockchain has emerged as a transformative technology for streamlining and increasing transparency in these systems; however, issues regarding latency, scalability, throughput, and security often restrict its successful implementation. Hyperledger Fabric, a prominent private blockchain architecture, has gained significant traction in the healthcare supply chain due to its modular design, privacy features, and high transaction efficiency. To further address the challenge of scalability, sharding is introduced as a technique to divide the network into smaller, evenly distributed segments called shards. These shards process transactions in parallel rather than taxing the entire network. This review paper offers an overview of existing studies by examining the integration of Hyperledger Fabric and sharding within the healthcare supply chain. It addresses a critical gap in the literature where investigations into scalability and security are dispersed.
When the power-frequency short-circuit current flows through the armor rods segment on an overhead ground wire (OGW), the OGW at the segment may experience fracture failure due to high temperatures. Consequently, it is necessary to optimize the structural configuration of the armor rods segment. Based on the structural characteristics of the conventional armor rods segment, this paper proposes a stepped-type armor rods segment structure. First, an electromagnetic–thermal coupling simulation model for both types of armor rods segment is constructed, in which the conductor length is determined by the boundary conditions of both the electromagnetic field and the thermal field. The current density distribution and transient temperature distribution under power-frequency short-circuit current are analyzed using the simulation model. Subsequently, based on the simulation results, an evaluation method for the mechanical performance of the OGW considering non-uniform temperature distribution is proposed. This method is employed to compare the high-temperature mechanical properties of the OGW at the two types of ends. Finally, a transient temperature rise experiment is designed to validate the accuracy of the simulation model. The research results show that the simulation model has sufficient accuracy, with an error of no more than 6%. Compared with the conventional armor rods segment, the stepped-type structure effectively avoids the concentration of high-temperature zones. Under identical conditions, the mechanical load-bearing capacity of the OGW at the stepped-type end is higher than that at the conventional end, which can help prevent high-temperature fracture of the OGW to a certain extent.
Semantic search over domain-specific corpora requires an effective embedding model and infrastructure. Elasticsearch’s native semantic_text field and ELSER sparse-vector inference require a commercial Enterprise subscription, inaccessible to most academic institutions. This paper documents the licensing barrier and presents a reproducible manual pipeline achieving comparable semantic search with free Elasticsearch components and open-source small language models (SLMs), on a four-node Raspberry Pi 4 cluster (8 GB RAM, three-node Elasticsearch) over 1088 USGS documents. Five models were evaluated —ELSER v2, .multilingual-e5-small, all-MiniLM-L12-v2, all-mpnet-base-v2, and msmarco-MiniLM-L12-cos-v5—from 35 screened candidates, plus a BM25 lexical baseline. Available process memory, not compute, is the binding constraint: Elasticsearch’s footprint consumes 5–6 GB of the 8 GB. The three 384-dimensional models reindexed the corpus in 2.1–2.2 h; the 768-dimensional all-mpnet-base-v2 took 10.1 h. On-disk size is unreliable for provisioning: .multilingual-e5-small expands from 1.4 GB on disk to 3.5 GB at runtime (2.5×). Retrieval quality was assessed with Precision@10, MRR, MAP@10, and nDCG@10 over 15 human-judged queries rather than raw similarity scores; embedding-based retrieval outperforms BM25, reaching significance for two of five configurations. Msmarco-MiniLM-L12-cos-v5 offers the strongest quality-per-resource trade-off among the 384-dimensional models for 8 GB ARM deployments. Pipeline and configuration artefacts are documented and reproducible.
Reconfigurable intelligent surfaces (RISs) are emerging as a key enabling technology to engineer the wireless propagation environment in 5G/6G systems. This paper presents the design, characterization, and control of a high-resolution RIS targeting slowly time-varying scenarios, in which fine-grained and stable phase control is more valuable than fast reconfiguration. Starting from the OpenRIS unit-cell layout, the cell is re-optimized for single-polarization operation and continuous phase tuning through a single varactor diode, exploiting the full tuning range enabled by a high-resolution DAC infrastructure rather than multi-bit quantization. The unit cell is analyzed via full-wave 3D FEM simulation in COMSOL Multiphysics and optimized to maximize the reflection-phase excursion while limiting amplitude modulation across the 5G N78 band (3.60–3.78 GHz). The design is experimentally validated in a WR-284 waveguide fixture, exhibiting a phase excursion approaching the full 360∘ near resonance, with an amplitude variation below 1 dB over the bias sweep at any given frequency, while the average reflection level decreases by about 2 dB from the center to the upper band edge. A fifth-order polynomial phase–voltage calibration feeds a lookup table driving a layered control system based on a Python HMI, a server, and STM32-driven 16-bit DACs. Experimental measurements confirm that the control chain delivers the commanded bias voltages to the addressed unit cells within measurement uncertainty; the array-level beamforming is assessed at simulation level under idealized (unit-magnitude) assumptions, while the experimental characterization of the assembled surface is left to future work.
In SRAM-based compute-in-memory (CIM), read-bitline (RBL) charging and discharging depend on the physical bit pattern stored in the memory array, so the energy-relevant code statistic should be defined with respect to the actual read-port polarity. This paper presents a read-polarity-aware row-wise offset-encoding method for W4A8 INT4 weights. In the evaluated Q-sensed 8T topology, the stored logical one is the discharge-active state; hence, the topology-specific read-active density equals the stored-one fraction. Under the exact whole-row INT4-feasibility protocol, a nonzero row offset is accepted only when every translated valid signed-INT4 code remains within [−8, 7]; no clipping, saturation, wraparound, or remapping is permitted, and zero offset remains the fallback. The complete software evaluation covers 286 quantized modules, 579,464 physical 16 × 16 tile positions, and 147,156,296 quantized weights across ResNet-18, MobileNetV3-Small, and SmolLM2-135M. Circuit re-validation uses 300 independent tile-policy samples, 1200 matched baseline-selected bitplane pairs, and 2400 successfully completed transistor-level Spectre simulations. The balanced circuit population yields an aggregate local SRAM readout-energy reduction of 5.03%, with a sample-cluster bootstrap 95% confidence interval of 4.12–6.02%. After four-bitplane aggregation, relative read-active-density reduction and local SRAM readout-energy reduction exhibit Pearson r = 0.81 and Spearman ρ = 0.75, indicating a substantial but imperfect relationship. The directly validated no-offset, positive-offset, and signed-offset policies preserve the corresponding model-level Top-1 accuracy or perplexity. Proposal-specific digital overheads and metadata-storage capacity are quantified separately, whereas representative physical SRAM/ROM metadata-access energy remains uncharacterized. Accordingly, the measured energy benefit is limited to local SRAM readout; a net energy reduction at the complete CIM-macro or system level, robustness across all evaluated PVT conditions, and robustness to process mismatch are not established by the present evidence.
Millimeter-wave (mmWave) radar-based gesture recognition has attracted increasing attention for real-time human–computer interaction owing to its robustness to illumination changes, privacy-preserving sensing capability, and suitability for embedded deployment. However, existing single-stream models often couple heterogeneous point-cloud and temporal statistical features, which may limit their ability to capture fine-grained motion patterns and key action frames. To address this problem, this paper proposes a dual-stream long short-term memory network (LSTM) and a bidirectional gated recurrent unit (BiGRU) with an attention mechanism (Attention-BiGRU) network for mmWave radar-based hand gesture recognition, termed as DSTG-Net. In the proposed DSTG-Net framework, an LSTM branch is used to process radar point-cloud sequences and extract fine-grained spatio-temporal features, while an Attention-BiGRU branch models global motion trends from statistical and temporal-difference features. The attention mechanism in the Attention-BiGRU branch is introduced to emphasize discriminative frames during gesture transitions, and the complementary features from the two branches are fused through feature concatenation for final classification. The proposed method is evaluated on a public mmWave radar gesture dataset to verify its recognition performance, and an additional self-built near-field dataset is used to test its effectiveness under a constrained acquisition scene. The proposed method achieves recognition accuracies of 97.40% and 98.75% on the two datasets, respectively, outperforming several baseline models. The Raspberry Pi-based implementation with a TI IWR1642 radar confirmed the functional feasibility of the proposed pipeline.
