Smart Image Based Technology and Deep Learning for Tunnel Inspection and Asset Management
2
Citation
0
Reference
20
Related Paper
Citation Trend
Abstract:
Tunnel inspection and asset management is typically a labour-intensive process where engineering judgement and experience is heavily relied upon to identify and assess tunnel condition over kilometres of homogeneous structures. Novel work flows and digital applications have been developed by the authors to create various smart image-based inspection and analysis tools that reduce the potential subjectivity and inconsistency of these inspections. This has resulted in significant improvements to existing tunnel inspection practices and structural health assessment. Current advances in image capture technology and computational processing power has enabled high integrity data to be easily captured, visualised and analysed. The work flows and tools developed take advantage of existing low-cost image capture hardware, open-source processing software and couples this with the creation of unique machine learning algorithms and analytics. Core innovations include: (i) use of low-cost photographic equipment for high quality imagery capture (ii) use of automated inspection vehicles for data capture (iii) Deep learning for automatic defect object recognition and defect classification (iv) Creation of immersive dashboards and 3D visualisations. This results in a suite of image based service offerings and deliverables, relevant to specific tunnel engineering issues and asset management aims. Thanks to deep learning, defect detection and asset condition metrics are automatically created, enabling: (i) the tunnel owner to gain greater insights into their asset resilience and operations, (ii) the tunnel engineer to focus on key issues aided by machine learning.Cite
Cite
Citations (19)
Inspection is crucial to the management of ageing infrastructure. Visual information on structures is regularly collected but very little work exists on its organised and quantitative analysis, even though image processing can significantly enhance these inspection processes and transfer real financial and safety benefits to the managers, owners and users. Additionally, new opportunities exist in the fast evolving sectors of wind and wave energy to add value to image-based inspection techniques.
This book is a first for structural engineers and inspectors who wish to harness the full potential of cameras as an inspection tool. It is particularly directed to the inspection of offshore and marine structures and the application of image-based methods in underwater inspections. It outlines a set of best practice guidelines for obtaining imagery, then the fundamentals of image processing are covered along with several image processing techniques which can be used to assess multiple damage forms: crack detection, corrosion detection, and depth analysis of marine growth on offshore structures. The book provides benchmark performance measures for these techniques under various visibility conditions using an image repository which will help inspectors to envisage the effectiveness of the techniques when applied. MATLAB® scripts and access to the underwater image repository are included so readers can run these techniques themselves.
Practising engineers and managers of infrastructure assets are guided in image processing based inspection. Researchers can use this book as a primer, and it also suits advanced graduate courses in infrastructure management or on applied image processing.
Cite
Citations (2)
Digital Twin of an infrastructure is a living digital simulation that brings all the data and models together and updates itself from multiple sources to represent its physical counterpart. The primary focal point of the present study is to propose a framework for Digital Twin of infrastructure and to demonstrate it in the context of a next-generation condition assessment method. The proposed framework is based on the optimized integration of: (1) Structural Inspection: Autonomous Data Collection using drones to minimize intrusion on the transport flow, cover large areas in a minimum of time, access to hard-to-reach areas and minimize exposure to safety hazards for inspectors and users; (2) Damage Quantification: Automated Data Interpretation using data-driven techniques to detect and quantify geometrical and visual anomalies. e.g. cracking and spalling, on the surface and sub-surface of concrete infrastructure; and (3) Performance Prediction: Advanced Structural Simulation combined with physics-based deterioration models to calculate structural performance. The outcome of the study is expected to radically transform the current practices by leveraging drones for inspection, data-driven models for damage quantification, and physics-based models for performance prediction, all seamlessly connected to a living simulation platform “Digital Twin†which updates itself after each inspection round.
Cite
Citations (9)
Cite
Citations (0)
Construction site safety
Cite
Citations (8)
Abstract. Civil infrastructures, such as tunnels and bridges, are directly related to the overall economic and demographic growth of countries. The aging of these infrastructures increases the probability of catastrophic failures that results in loss of lives and high repair costs; all over the world, these factors drive the need for advanced infrastructure monitoring systems. For these reasons, in the last years, different types of devices and innovative infrastructure monitoring techniques have been investigated to automate the process and overcome the main limitation of standard visual inspections that are used nowadays. This paper presents some preliminary findings of an ongoing research project, named TACK, that combines advanced deep learning techniques and innovative photogrammetric algorithms to develop a monitoring system. Specifically, the project focuses on the development of an automatic procedure for crack detection and measurement using images of tunnels and bridges acquired with a mobile mapping system. In this paper, some preliminary results are shown to investigate the potential of a deep learning algorithm in detecting cracks occurred in concrete material. The model is a CNN (Convolutional Neural Network) based on the U-Net architecture; in this study, we tested the transferability of the model that has been trained on a small available labeled dataset and tested on a large set of images acquired using a customized mobile mapping system. The results have shown that it is possible to effectively detect cracks in unseen imagery and that the primary source of errors is the false positive detection of crack-like objects (i.e., contact wires, cables and tile borders).
Bridge (graph theory)
Transferability
Cite
Citations (10)
When it comes to addressing the safety/security related needs at different production/construction sites, accurate detection of the presence of workers, vehicles, equipment important and formed an integral part of computer vision-based surveillance systems (CVSS). Traditional CVSS systems focus on the use of different computer vision and pattern recognition algorithms overly reliant on manual extraction of features and small datasets, limiting their usage because of low accuracy, need for expert knowledge and high computational costs. The main objective of this paper is to provide decision makers at sites with a practical yet comprehensive deep learning and IoT based solution to tackle various computer vision related problems such as scene classification, object detection in scenes, semantic segmentation, scene captioning etc. Our overarching goal is to address the central question of What is happening at this site and where is it happening in an automated fashion minimizing the need for human resources dedicated to surveillance. We developed Deep ExxonMobil Eye for Video Analysis (DEEVA) package to handle scene classification, object detection, semantic segmentation and captioning of scenes in a hierarchical approach. The results reveal that transfer learning with the RetinaNet object detector is able to detect the presence of workers, different types of vehicles/construction equipment, safety related objects at a high level of accuracy (above 90%). With the help of deep learning to automatically extract features and IoT technology to automatic capture, transfer and process vast amount of realtime images, this framework is an important step towards the development of intelligent surveillance systems aimed at addressing myriads of open ended problems in the realm of security/safety monitoring, productivity assessments and future decision making.
Closed captioning
Transfer of learning
Machine Vision
Cite
Citations (2)
Integrating design and operating envelopes of an asset with live operations data allows operators to see exceedances in real time and precisely understand the health of their asset. This technology is a direct stem of the rapidly growing digitalisation philosophy of many operators, whereby data that have traditionally been locked in PDF documents, Excel spreadsheets and other silos can be digitised, structured and connected to many different end points. For asset integrity monitoring, this means digitising and integrating design envelopes, data historians, enterprise resource planning systems and calculation tools with a unified front-end visualisation, to display what is really happening on an asset. By integrating and visualising these data sources, it is possible to create a single location for monitoring the health of all operator assets and generating reports for continued assessment. Once established, the platform provides a base for deployment of other intelligent toolkits, such as pattern recognition for valve signature monitoring, or predictive modelling for corrosion management over the life of field. This extended abstract uses a case study to discuss the benefits of standardising the visualisation landscape for monitoring the health of assets, the improvements and efficiencies delivered to the integrity monitoring process, the ability to highlight flaws in the design envelopes, and the engineering required to ensure that the right instrument is monitoring the right equipment.
Corrosion monitoring
Cite
Citations (1)
Abstract. In this work, it is examined the 2D recognition and 3D modelling of concrete tunnel cracks, through visual cues. At the time being, the structural integrity inspection of large-scale infrastructures is mainly performed through visual observations by human inspectors, who identify structural defects, rate them and, then, categorize their severity. The described approach targets at minimum human intervention, for autonomous inspection of civil infrastructures. The shortfalls of existing approaches in crack assessment are being addressed by proposing a novel detection scheme. Although efforts have been made in the field, synergies among proposed techniques are still missing. The holistic approach of this paper exploits the state of the art techniques of pattern recognition and stereo-matching, in order to build accurate 3D crack models. The innovation lies in the hybrid approach for the CNN detector initialization, and the use of the modified census transformation for stereo matching along with a binary fusion of two state-of-the-art optimization schemes. The described approach manages to deal with images of harsh radiometry, along with severe radiometric differences in the stereo pair. The effectiveness of this workflow is evaluated on a real dataset gathered in highway and railway tunnels. What is promising is that the computer vision workflow described in this work can be transferred, with adaptations of course, to other infrastructure such as pipelines, bridges and large industrial facilities that are in the need of continuous state assessment during their operational life cycle.
Initialization
Cite
Citations (9)
Facade
Identification
Building Automation
Cite
Citations (6)