Open-source and free software for digital image correlation (DIC) and digital volume correlation (DVC) have developed rapidly in recent years. However, the functionality, availability, and measurement capabilities of existing DIC/DVC freeware remain insufficient to adequately support scientific research and engineering applications. This paper presents OpenCorr GUI (version 3), a standalone software developed based on OpenCorr library, which provides comprehensive functionalities for deformation measurements using 2D-DIC, Stereo-DIC, and DVC, as well as a user-friendly graphical user interface (GUI) for parameter configuration and result visualization. Moreover, a regional fitting strategy is proposed and integrated into the software to enhance the capability of DIC/DVC for measuring complex large deformation. The software provides a specific approach for strain calculation for Stereo-DIC, which is not included or correctly implemented in most existing DIC freeware. The functions of OpenCorr GUI are demonstrated through a series of simulated and real experimental cases, including large tensile deformation measurement in 2D-DIC, inhomogeneous compressive deformation measurement in DVC, complex surface topography reconstruction and strain measurement on the curved surface of an inflated balloon in Stereo-DIC. The results show that the software provides accurate and robust measurements, while offering comprehensive features and a good user experience.
Physics-informed and hybrid machine learning approaches are emerging as critical enablers of robust and autonomous friction stir welding (FSW) systems. While conventional finite element models provide physical interpretability but suffer from high computational cost, and purely data-driven machine learning models offer predictive speed but lack generalization and explainability, neither approach alone is sufficient for safety–critical and industrial-scale welding applications. This review critically examines the evolution from empirical and physics-based modeling toward hybrid and physics-informed artificial intelligence frameworks that integrate thermomechanical governing equations with data-driven learning architectures. The paper synthesizes recent advances in supervised and unsupervised learning for weld quality prediction, defect detection, tool wear monitoring, and microstructure estimation, and highlights their limitations in transferability, data dependency, and certification readiness. Particular emphasis is placed on physics-informed neural networks (PINNs), hybrid finite element-machine learning models, reduced-order surrogate modeling, and multi-fidelity data fusion strategies that embed conservation laws, constitutive behavior, and process constraints directly into learning algorithms. These approaches demonstrate improved robustness, reduced training data requirements, enhanced extrapolation capability, and greater interpretability compared to purely black-box models. The review further examines explainable artificial intelligence (XAI) techniques for linking process parameters to microstructural evolution and mechanical performance and discusses the integration of hybrid models into real-time control architectures, reinforcement learning frameworks, and digital twin systems. Key challenges including data standardization, model generalization across alloys and joint geometries, computational efficiency, cybersecurity, and industrial validation are critically analyzed. By unifying thermomechanical physics with adaptive learning, physics-informed modeling provides a pathway toward trustworthy, scalable, and self-optimizing FSW systems. The review concludes by outlining a roadmap for autonomous intelligent welding platforms capable of supporting next-generation aerospace and automotive manufacturing within Industry 4.0 ecosystems.
Wireless Sensor Networks (WSNs) are pivotal to Internet of Things (IoT) infrastructures, facilitating real-time environmental, industrial, and agricultural tracking. However, the limited energy and computation capabilities of sensor nodes (SNs) severely restrict network lifetime and reliability. This paper proposes an Event Preserving Adaptive Aggregation (EP-Agg) protocol integrated with an Energy Weighted Minimum Spanning Tree (EW-MST) clustering combined with a Memetic Algorithm (MA) based Cluster Head (CH) optimization for energy-efficient WSN data transmission. The proposed approach employs adaptive volatility-driven segmentation to compress redundant sensor readings while preserving significant events. EW-MST clustering combined with an MA ensures energy-balanced topology and optimized CH selection. In terms of network stability, the designed methodology outperforms Hybrid Particle Swarm Optimization −Improved Low-Energy Adaptive Clustering Hierarchy (HPSO-ILEACH) by 2.2 times, Double Exponential Adaptive Inertia (DEAI) by 4 times, and Adaptive Hybrid Dragonfly Firefly (AHDF) by an average of 42 times across the evaluated network densities, with improvements ranging from 25.8 times to 70.8 times depending on node density, according to MATLAB-based simulations. Similarly, the proposed method outperforms HPSO-ILEACH by 1.2 times, DEAI by 1.7 times, and AHDF by 2.7 times in terms of network lifetime. Moreover, this algorithm achieves a residual energy efficiency of 44% with respect to the initial network energy, which is 15% more than HPSO-ILEACH, 22% more than DEAI, and 33% more than AHDF, showing its enhanced energy efficiency. Additional event-fidelity evaluation using multi-sensor data demonstrates that EP-Agg provides improved preservation of transient events compared with conventional Moving Average and Piecewise Aggregate Approximation (PAA).
The accuracy of long-term machine vision-based structural displacement monitoring is critically compromised by temperature-induced image drift. This study investigates the underlying mechanisms by developing a projection-error model and applying variance-based global sensitivity analysis (GSA) to evaluate the contributions of camera parameters to the image drift. Based on the GSA identification of dominant parameters, a simplified, and physically interpretable error model is formulated. Principal-point shifts and rotations about the camera’s lateral axes (Euler angles) are identified as the dominant error sources induced. A key finding is the focal-length-dependent amplification of rotational errors, scaling linearly with focal length, whereas principal-point effects remain independent of object distance. This mechanism, which has not been previously quantified or validated, explains why rotations become significant in long‑range monitoring despite their small magnitude.. Field experiments and a dam monitoring case validate the model, demonstrating that drift observed on distant stationary targets is highly correlated with measurement error. This indicates that such stationary features can potentially serve as in-situ references for error compensation. The work establishes a theoretical basis and offers practical guidance for measurement error control in videogrammetric systems, contributing to more accurate long-term vision-based displacement monitoring.
