Regular inspections of civil structures and infrastructure, performed by professional inspectors, are costly and demanding in terms of time and safety requirements. Additionally, the outcome of inspections can be subjective and inaccurate as they rely on the inspector’s expertise. To address these challenges, autonomous inspection systems offer a promising alternative. However, existing robotic inspection systems often lack adaptive positioning capabilities and integrated crack labelling, limiting detection accuracy and their contribution to long-term dataset improvement. This study introduces a fully autonomous framework that combines real-time crack detection with adaptive pose adjustment, automated recording and labelling of defects, and integration of RGB-D and LiDAR sensing for precise navigation. Damage detection is performed using YOLOv5, a widely used detection model, which analyzes the RGB image stream to detect cracks and generates labels for dataset creation. The robot autonomously adjusts its position based on confidence feedback from the detection algorithm, optimizing its vantage point for improved detection accuracy. Experiment inspections showed an average confidence gain of 18% (exceeding 20% for certain crack types), a reduction in size estimation error from 23.31% to 10.09%, and a decrease in the detection failure rate from 20% to 6.66%. While quantitative validation during field testing proved challenging due to dynamic environmental conditions, qualitative observations aligned with these trends, suggesting its potential to reduce manual intervention in inspections. Moreover, the system enables automated recording and labeling of detected cracks, contributing to the continuous improvement of machine learning models for structural health monitoring.

Surrogate models have gained widespread popularity for their effectiveness in replacing computationally expensive numerical analyses, particularly in scenarios such as design optimization procedures, requiring hundreds or thousands of simulations. While one-shot sampling methods—where all samples are generated in a single stage without prior knowledge of the required sample size—are commonly adopted in the creation of surrogate models, these methods face significant limitations. Given that the characteristics of the underlying system are generally unknown prior to training, adopting one-shot sampling can lead to suboptimal model performance or unnecessary computational costs, especially in complex or high-dimensional problems. This paper addresses these challenges by proposing a novel, model-independent adaptive sampling approach with batch selection, termed Cross-Validation Batch Adaptive Sampling for High-Efficiency Surrogates (CV-BASHES). CV-BASHES is first validated using two analytical functions to explore its flexibility and accuracy under different configurations, confirming its robustness. Comparative studies on the same functions with two state-of-the-art methods, maximum projection (MaxPro) and scalable adaptive sampling (SAS), demonstrate the superior accuracy and robustness of CV-BASHES. Its applicability is further demonstrated through a geotechnical application, where CV-BASHES is used to develop a surrogate model to predict the horizontal deformation of a diaphragm wall supporting a deep excavation. Results show that CV-BASHES efficiently selects training samples, reducing the dataset size while maintaining high surrogate accuracy. By offering more efficient sampling strategies, CV-BASHES streamlines and enhances the process of creating machine learning models as surrogates for tackling complex problems in general engineering disciplines.

This paper presents a reliability-constrained Bayesian optimization framework for structural design under uncertainty, addressing challenges in stochastic optimization where the objectives and constraints are defined implicitly by potentially expensive numerical models. Our approach explicitly accounts for parameter uncertainty using results from Bayesian quadrature for uncertainty propagation in Gaussian process surrogate models. The method accommodates arbitrary probability distributions and employs gradient-based optimization for acquisition function maximization, strategically selecting sample points to minimize numerical model evaluations. We demonstrate our algorithm’s superior performance over random search and conventional Bayesian optimization through both an analytical test function and a prestressed tie-beam design case study, showing its practical applicability to structural optimization problems.

Addressing and predicting degenerative phenomena in domains such as health care and engineering, two fundamental fields of vital importance for society, offers valuable insights into early warning steps and critical event forecasting, leading to far-reaching implications for safety and resource allocation. By harnessing the power of data-driven insights, prognostics becomes the principal component of predicting such phenomena. Developing clustering techniques as feature extractors acts as an intermediate step between the raw incoming data and prognostics and provides the opportunity to unveil hidden relationships within complex datasets. However, when limited, noisy, and multimodal data are available in a label-free format, extensive preprocessing, and unreliable, complicated models are required for extracting meaningful features. This prohibits the development of adaptable methods in diverse domains that are in favor of robustness and interpretability. In this regard, this study introduces a novel unsupervised deep clustering model for feature extraction in degenerative phenomena. The model innovatively extracts prognostic-related features from raw data via clustering analysis, characterized by an increasing monotonic behavior representing system deterioration. This monotonicity is partial rather than complete, to incorporate the potential occurrence of oscillations in the degradation trajectory of the system or noise-related data, reflecting real-world scenarios. Its performance, robustness, generalizability, and interpretability are evaluated across diverse domains utilizing three datasets from health care and engineering featuring limited, noisy, high-dimensional, and multimodal raw signals. Results show that the model extracts meaningful prognostic-related features in both domains and all datasets, without a significant alteration in its architecture and independently of the chosen prognostic algorithm.