Autonomous and AI-Driven Inspection and Damage Assessment of Infrastructure

Aging infrastructure, increasing environmental demands, and the growing impact of climate change are placing unprecedented pressure on civil structures worldwide. Many structures are approaching or exceeding their design life, while exposure to extreme events accelerates deterioration. At the same time, rapid advances in computer vision, robotics, and machine learning are transforming the way physical assets can be monitored and assessed. Traditional visual inspections, although widely practiced, remain subjective, labor-intensive, and difficult to document. Digital inspection approaches offer a promising alternative by enabling objective, data-driven, and scalable damage documentation. These technologies create opportunities to improve safety, efficiency, and decision-making across diverse infrastructure systems, from new assets to historical structures.This Research Topic aims to develop a systematic and end-to-end perspective on digital infrastructure inspection.1. Data collectionAdvances in UAVs, robotic platforms, photogrammetry, and multi-sensor systems enable large-scale, high-resolution data acquisition across above-water and underwater environments. Contributions are encouraged on optimization, automation, localization in GPS and GPS-denied conditions, data quality control, and repeatable inspection protocols.2. Data processing and analysisRobust AI and computer vision models are required for reliable damage detection, segmentation, and quantification. Particular emphasis should be placed on uncertainty quantification and generalization across structure types, and validation against experimental or field data.Contributions addressing the creation, sharing, and accessibility of high-quality ground-truth datasets, including openly available reference data to support benchmarking and validation of AI models, are also strongly encouraged.3. Automated damage interpretation and performance evaluationDigital inspection generates large volumes of detailed damage data. However, detection alone is insufficient. This Research Topic seeks contributions that move toward automated interpretation of detected damage, linking observed defects to durability, service life, and structural performance. Methods that integrate mechanics-based models, probabilistic approaches, and data-driven techniques to assess the influence of damage on load-bearing capacity and degradation processes are particularly encouraged.4. Data management, integration, and visualizationEfficient long-term asset management relies on scalable storage and interoperability standards. This includes reconstructing models from raw data and integrating them into BIM environments or digital twins. We prioritize advanced 3D visualization and decision-support systems to ensure that complex inspection data is translated directly into actionable engineering insights.Approaches that use browsable 3D models as living information environments, where inspection data can be progressively updated to track infrastructure deterioration over time, are also of strong interest.This Research Topic welcomes contributions addressing digital and AI-driven approaches for structural inspection and damage assessment. Topics of interest include, but are not limited to:● UAVs, robotics, and automation for infrastructure data acquisition, including above-water and underwater inspection systems; autonomous and semi-autonomous platforms operating in both GPS-available and GPS-denied environments ● Computer vision and deep learning for damage detection and segmentation● Uncertainty quantification and reliability assessment in AI-based predictions● Development of ground-truth datasets to support validation of AI-based inspection models● Data storage, interoperability, model reconstruction, and integration with existing BIM or digital twins, including automated scan-to-BIM workflows integrating multi-sensor data● Visualization techniques for large-scale 3D inspection data, including browsable 3D environments supporting progressive updating● Case studies on reinforced concrete, steel, and historical masonry constructionsWe encourage original research articles, review papers, and case studies that demonstrate methodological advances, validation on real structures, the development of openly accessible reference datasets, or integrated inspection frameworks.Dr. Amir Rezaie is a founder and shareholder at SwissInspect. He is currently the CEO of SwissInspect. The other Topic Editors declare no competing interests with regard to the Research Topic subject. Topic editors: Amir Rezaie, Chiara Casarotti, Florian Noichl, Qianqing Wang, Liangfu Ge Source: https://www.frontiersin.org/research-topics/79194/autonomous-and-ai-driven-inspection-and-damage-assessment-of-infrastructure