Experienced Inspectors Research Articles

Weld Defect Downscaling Recognition for Phased Array Ultrasonic Data Based on Semantic Segmentation

Phased array ultrasonic testing (PAUT) is one of the most advanced ultrasonic inspection methods, providing inspection data in image sequences with increased resolution and coverage, consequently enhancing inspection efficiency. Due to the inherently abstract nature of ultrasound images, the analysis of PAUT weld data continues to rely on the expertise of experienced inspectors, particularly in the context of shipbuilding welds, which exhibit characteristics such as arbitrary lengths, varying thicknesses, and various weld methods, particularly when associated with hybrid laser-arc welding involving higher caps and narrower fill widths. This paper applies a semantic segmentation architecture to PAUT data for hybrid laser-arc welding to segment defects automatically. A downscaling PAUT image dataset incorporating PAUT imaging principles with scan processes is established. The proposed architecture adopts an encoder-decoder architecture and segments defects downscaling, enabling direct localization and sizing defects in long PAUT data. Meanwhile, it handles PAUT image sequences of arbitrary length, eliminating the need for horizontal scale normalization while preserving the integrity of defect shapes. The proposed architecture was validated in the dataset to demonstrate the superiority of the proposed method in locating and sizing defects in PAUT weld data, offering a promising automated solution to a challenge in ultrasonic inspection methodologies.

Artificial intelligence-aided detection of rail defects based on ultrasonic imaging data

Railroads are one of society’s fundamental infrastructures, facilitating the transportation of passengers and goods over vast distances. Rail status data is immensely important for ensuring the safe and efficient operation of railroad networks. However, analyzing ultrasonic inspection data is a labor-intensive process and relies heavily on the expertise of experienced inspectors. To detect internal defects of the rail accurately and automatically, this paper proposes a customized image recognition method based on a convolutional neural network with limited B-scan rail image data collected within the industry. The proposed method uses EfficientNet-b7 as the backbone network to fully extract the B-scan rail image data features. With the help of transfer learning and data augmentation techniques, the backbone network is substantially enhanced so that it can understand high-level features of the object without being trained with large-scale B-scan image data. We establish a real-world internal rail defect dataset with 280 B-scan images and test our proposed method. The results reveal that the highest accuracy of the other mainstream CNN-based methods is 76.25% and the accuracy of the traditional method based on a support vector machine classifier trained with Tamura texture and LBP features is 60.00%. Our proposed EfficientNet-b7 model classifies rail defect B-scan images with an accuracy of 85.00%, precision of 84.71%, and recall of 85.00%. Compared to other rail internal defect detection methods, this method is more accurate. With the help of transfer learning and data augmentation, our proposed method achieves better performance and requires less data.

Automated Pavement Construction Inspection Using Uncrewed Aerial Systems (UAS)—Hot Mixed Asphalt (HMA) Temperature Segregation

Temperature segregation in Hot Mixed Asphalt (HMA) pavement construction leads to performance problems, such as reduced fatigue life. During construction, Quality Assurance (QA) inspection procedures are required to evaluate the pavement condition and detect the segregated areas. In traditional HMA highway construction inspection processes, temperature differences are investigated manually, by sampling the HMA behind the paver. In these processes, inspectors are required to work adjacent to traffic and alongside moving or backing equipment. These processes do not provide a complete temperature profile of the mat, endanger the inspectors’ safety, and require on-site experienced inspectors. An Uncrewed Aerial System (UAS) enables HMA pavement construction inspection to be conducted within a remote, non-destructive, safe, and efficient framework. The objective of this research is to design an automated UAS imaging workflow for HMA pavement construction inspection, mainly locating temperature segregation. The primary contribution of this paper is to provide Departments of Transportations (DOTs) and contractors with workflows for creating enhanced remote inspection procedures and detailed thermal profiles of the placed HMA mat. The application of the proposed workflow is illustrated using an HMA construction project in Michigan.

Approach for Estimating Inspection Staffing Needs for Highway Construction Projects

Construction inspectors are the primary workforce in the highway construction industry who ensure that the work on infrastructure projects meets or exceeds the quality standards set. However, there is a growing concern in state departments of transportation (DOTs) about the inspection process because of the declining availability of experienced inspectors. This study proposes an approach for estimating inspection staffing needs for highway construction projects. Data from 157 construction projects completed between 2017 and 2021 were used to evaluate project inspection staff size based on the project type, cost, and duration. A total of 35 core inspection items were identified by carrying out a content analysis of construction inspection documents. Then, a focus group of 18 experts from the Kansas DOT was conducted to verify the list of 35 cost inspection items and evaluate the number and experience of inspectors required for each item. The four project types included in the analysis are as follows: grading/surfacing; sealing; surface recycling and overlay; and bridge repair. The analysis showed that the inspection staffing needs depend on project type, duration, and size. The grading/surfacing and bridge repair projects require more inspection staff than the sealing, and surface recycling and overlay projects. The result of a Spearman’s correlation analysis indicated that there is a significant association between the risk level of an inspection item and the number and experience of inspectors required for that item. The findings from this study may help state DOTs identify their needs for inspection staff and improve the quality of highway construction projects.

Experience-based systematic construction hazard warning

PurposeHazard warning schemes provide efficient hazard recognition and promote project safety. Nevertheless, these schemes perform poorly because the warning information is calibrated for individual characters and is not prioritized for the entire system. This study proposes a hazard warning scheme that prioritizes hazard characters from the inspection process based on the inspectors' experience.Design/methodology/approachFirst, hazard descriptions were decomposed into their characters, forming a double-layer network. Second, warning schemes based on cascading effects were proposed. Third, character-based warning schemes were simulated for various experiences.FindingsThe results show that when a specific hazard is detected, the degree centrality is the most effective parameter for prioritization, and hazard characters should be prioritized based on betweenness centrality for experienced inspectors, whereas degree centrality is preferred for novice inspectors.Originality/valueThe warning scheme theoretically supplements the information-processing theory in construction hazard warnings and provides a practical warning scheme with priority for the development of automated hazard navigation systems.

A defect detection method for topological phononic materials based on few-shot learning

Topological phononic materials have been widely used in many fields, such as topological antennas, asymmetric waveguides, and noise insulation. However, due to the limitations of the manufacturing process, topological protection is vulnerable to some severe defects that may affect the application effect. Therefore, the quality inspection of topological materials is essential to ensure reliable results. Due to the low contrast and irregularity of defects and the similarity of topological phononics, they are difficult to recognize by traditional image processing algorithms, so manual detection is still mainstream at present. But manual detection requires experienced inspectors, which is expensive and time-consuming. In addition, topological materials are expensive to produce, and there is no large publicly available dataset, but deep learning usually relies on large datasets for training. To solve the above problems, we propose an automatic deep learning topology structure defect detection method (ADLTSDM) in this work, which could classify not only the structure of topological materials but also detect the defects of topological phononics based on a small dataset. ADLTSDM exploits the prior knowledge of the topological material structure and achieves an augmentation factor of more than 100 times through the random and fixed interval screenshot algorithm, thus enabling the training of deep neural networks with only two raw data. For defect detection, ADLTSDM has an accuracy of more than 97% and improves detection speed by more than 38% compared with manual detection. For structure classification, ADLTSDM can achieve an accuracy of over 99% and seven times faster speed compared with manual classification. Besides, the detection standard of ADLTSDM is unified, so the accuracy will not be affected by the experience of the inspectors, which has more potential in high-throughput industrial applications.

Detection of Cytopathic Effects Induced by Influenza, Parainfluenza, and Enterovirus Using Deep Convolution Neural Network.

The isolation of a virus using cell culture to observe its cytopathic effects (CPEs) is the main method for identifying the viruses in clinical specimens. However, the observation of CPEs requires experienced inspectors and excessive time to inspect the cell morphology changes. In this study, we utilized artificial intelligence (AI) to improve the efficiency of virus identification. After some comparisons, we used ResNet-50 as a backbone with single and multi-task learning models to perform deep learning on the CPEs induced by influenza, enterovirus, and parainfluenza. The accuracies of the single and multi-task learning models were 97.78% and 98.25%, respectively. In addition, the multi-task learning model increased the accuracy of the single model from 95.79% to 97.13% when only a few data of the CPEs induced by parainfluenza were provided. We modified both models by inserting a multiplexer and de-multiplexer layer, respectively, to increase the correct rates for known cell lines. In conclusion, we provide a deep learning structure with ResNet-50 and the multi-task learning model and show an excellent performance in identifying virus-induced CPEs.

A Markov chain-based model for structural vulnerability assessmentof corrosion-damaged reinforced concrete bridges.

The deterioration and cracking of reinforced concrete (RC) bridges due to the chloride-induced corrosion of steel reinforcement is an inherently time-dependent stochastic phenomenon. In the current practice of bridge management systems, however, the determination of the condition states of deteriorated bridges is highly dependent on the opinion of experienced inspectors. Taking such complexity into account, the current paper presents a new stochastic predictive methodology using a non-homogeneous Markov process, which directly relates the visual inspection data (corrosion rate and crack widths) to the structural vulnerability of deteriorated concrete bridges. This methodology predicts the future condition of corrosion-induced damage (concrete cracking) by linking structural vulnerability analysis and a discrete-time Markov chain model. The application of the proposed methodology is demonstrated through a case-study corrosion-damaged RC bridge pier.This article is part of a discussion meeting issue ‘A cracking approach to inventing new tough materials: fracture stranger than friction’.

A semi-supervised self-training method to develop assistive intelligence for segmenting multiclass bridge elements from inspection videos

Bridge inspection is an important step in preserving and rehabilitating transportation infrastructure for extending their service lives. The advancement of mobile robotic technology allows the rapid collection of a large amount of inspection video data. However, the data are mainly the images of complex scenes, wherein a bridge of various structural elements mix with a cluttered background. Assisting bridge inspectors in extracting structural elements of bridges from the big complex video data, and sorting them out by classes, will prepare inspectors for the element-wise inspection to determine the condition of bridges. This article is motivated to develop an assistive intelligence model for segmenting multiclass bridge elements from the inspection videos captured by an aerial inspection platform. With a small initial training dataset labeled by inspectors, a Mask Region-based Convolutional Neural Network pre-trained on a large public dataset was transferred to the new task of multiclass bridge element segmentation. Besides, the temporal coherence analysis attempts to recover false negatives and identify the weakness that the neural network can learn to improve. Furthermore, a semi-supervised self-training method was developed to engage experienced inspectors in refining the network iteratively. Quantitative and qualitative results from evaluating the developed deep neural network demonstrate that the proposed method can utilize a small amount of time and guidance from experienced inspectors (3.58 h for labeling 66 images) to build the network of excellent performance (91.8% precision, 93.6% recall, and 92.7% f1-score). Importantly, the article illustrates an approach to leveraging the domain knowledge and experiences of bridge professionals into computational intelligence models to efficiently adapt the models to varied bridges in the National Bridge Inventory.

Automated Crack Detection and Damage Index Calculation for RC Structures Using Image Analysis and Fractal Dimension

AbstractDamage assessment in postearthquake reconnaissance of civil structures has traditionally relied upon the judgment of experienced site inspectors. In a step toward eliminating the need for o...

Developing scaled tools for residential and nursing home inspection: feasibility study

PurposeThis study explored the feasibility of developing scaled inspection tools for use during external inspection of health and social care facilities to give improved accuracy in identifying facilities “at risk”, a tool for risk-adjusted frequency of inspection, and greater consistency of judgements.Design/methodology/approachThis paper summarises the development through working groups and workshops involving 20 experienced inspectors (nurses and social workers) of the Regulation and Quality Improvement Authority who inspect the 206 nursing and 182 residential care homes in Northern Ireland. A brief evaluation survey, including response to a case vignette, gathered inspectors' views after using the tools for six months.FindingsEight two-dimensional Scaled Inspection Tools were created, each embodying a scale of performance (seriousness of risk issue) and a scale of the ability of the facility to manage that issue, each axis comprising four points. The Scaled Inspection Tools were used for on-site inspections during 2017–18. Evaluative comments were generally positive. The case vignette seemed to highlight greater risk aversion amongst newer inspectors.Research limitations/implicationsThe creation of scaled inspection tools adds credibility to the potential for developing risk-based governance in service regulation. Further testing of domains and their scope is required.Practical implicationsPrompts for each domain were found essential to guide inspectors. Despite the challenge of change, inspectors became enthusiastic about use for evaluating risks, and managers about improvements in consistency of inspection.Social implicationsKnowledge derived from statistical approaches needs to be incorporated into inspection and regulation, just as in other aspects of professional practice.Originality/valueScaled inspection tools, with two orthogonal axes corresponding to seriousness of risk and ability to manage the risk (inverse of likelihood of harm), proved acceptable and intuitive in use. The study gives credibility to the possibility of developing screening and surveillance approaches to risk-based governance in service regulation.

Augmenting a deep-learning algorithm with canal inspection knowledge for reliable water leak detection from multispectral satellite images

Maintenance planning of groundwater delivery infrastructure, such as canals, requires labor-intensive field inspection for properly allocating maintenance resources to sections of water infrastructure based on their deterioration conditions. Defective canal sections have cracks where the water delivery performance degrades. In practice, canals can be tens or even hundreds of miles long. Manual canal inspections could take weeks, while could hardly achieve comprehensive water leakage assessment. Another difficulty is that most cracks are developing under the water. Without drying up the canals, inspectors could not observe underwater conditions. They would have to assess visible parts of water facilities and environments (e.g., humidity changes and vegetation growths nearby) for prioritizing canal sections in terms of leaking risks. Even experienced inspectors need much time to complete a reliable canal condition assessment.This paper presents a deep-learning approach augmented by canal inspection knowledge to achieve automated and reliable water leak detection of canal sections from Landsat 8 satellite images. Such integration utilizes the domain knowledge of experienced inspectors in augmenting the deep-learning methods for more reliable image pattern classification that supports rapid canal condition assessment. Compared with machine learning algorithms trained by raw satellite images manually labeled as leaking, domain-knowledge-augmented deep learning algorithms use satellite image augmented by pixel-level land surface temperature (LST), fractional vegetation coverage (FVC) and Temperature Vegetation Dryness Index (TVDI) as training samples. Specifically, LST, FVC, and TVDI for each pixel are physical parameters derived from Landsat 8 satellite images by remote sensing methods. The “leaking” or “no-leaking” labels of the training samples are from the concrete surface inspection records collected during annual dry-ups of the canal from 2016 to 2019. Testing results on data sets collected for canals flowing through both urban and rural areas show that the proposed approach can achieve recall at 86%, precision at 86%, and accuracy at 85%. The precision, recall, and accuracy of the proposed approach are similar to a conventional deep learning algorithm that uses raw images for training while being more computationally efficient. The reason is that the new approach only processes three channels rather than the 11 channels in raw images. The authors also tested how different combinations of environmental features influence the performance of the algorithm. The results showed that two feature combinations: (LST, FVC) and (LST, FVC, TVDI) achieve the most robust performance in diverse geospatial environments.

A mobile vision inspection system for tiny defect detection on smooth car-body surfaces based on deep ensemble learning

Defect inspection of car-body surfaces is a crucial step for assessing the appearance quality of car coatings. However, some problems such as uncertain imaging conditions, complicated local areas of car-body surfaces and tiny defects with ambiguous edges and low contrast in large images make defect detection more challenging. In order to overcome these challenges, a novel mobile inspection system based on computer vision technology is developed to inspect tiny defects on smooth car-body surfaces automatically. In our system, a dedicated image acquisition module (IAM) with a high-resolution camera, plane light-emitting diode light source and laser rangefinder is designed to capture surface images over bright field illumination. A specific deep ensemble learning algorithm, named TinyDefectNet, is proposed to identify tiny defects in large images acquired from the IAM online. The experiment results indicate our inspection system is on par with the average performance of experienced inspectors but it is much faster than artificial visual inspection. The developed inspection system has been successfully deployed in practice at a car coating plant and it has made inspection more efficient.

Ontology-Based Knowledge Management System for Digital Highway Construction Inspection

Inspection is critical to ensuring the quality of infrastructure construction. In recent years, state highway agencies (SHAs) have been facing the challenge of a shortage of experienced inspectors due to retirements, workforce downsizing, and resignations to take jobs in the private sector. There is a critical need to retain and manage the accumulated construction inspection knowledge (what, when, and how to inspect) and integrate this knowledge into the construction business process. This paper presents an ontological approach to managing inspection knowledge and the development of a risk-based digital inspection system integrated with the construction process. The developed system consists of five knowledge compartments: inspection activities prioritized based on risk; activity-centered pay items; inspection check items; detailed checks that include priority, frequency, object, attribute, and acceptance criteria; and training materials (optional). This risk-based system is designed to align with the construction process by connecting an activity to a pay item and that pay item to a check item, resulting in an automatically generated comprehensive inspection checklist for the activity. The optional training materials are designed to train inspectors in real time on an as-needed basis. Since risk is embedded in every knowledge compartment, this system not only automates and retains the inspection knowledge but also enables SHAs to allocate their limited resources to the most critical items. This system is currently being implemented by the Indiana Department of Transportation as a digital inspection protocol integrated in its e-construction process.

Bridge Inspections Come of Age

Arising from the tragic collapse of a West Virginia bridge 50 years ago this month, the National Bridge Inspection Program has helped keep bridges in the United States safe for motorists for decades. The standards and procedures for bridge inspections have changed dramatically over the years, and technology is fueling even greater transformations. But experts say it will be some time before robots and drones take the place of highly trained, experienced inspectors.

Application of Neural Network in Shop Floor Quality Control in a Make to Order Business

A make to order business has to produce the products that are customized to the customer's current need. The customization can be realized by assembling different standard parts with various ‘configurations’. The oil field service industry is a typical example where most products produced are cylindrical assemblies made up of standard parts customized in their size, material specifications, coating specifications, and threading suited for the particular load rating and environment. As business cycles go up and down, hiring and firing of personnel is the routine of the day. Thus, it is very hard to keep experienced inspectors due to high turnover of the staff on shop floor and thus intensive endeavor to train the inspectors for the same recurrent problems of the same standard parts is required. This paper proposes a neural network model to help the industrial practitioners address such a concern. The neural network is trained with ample ‘judgment calls’ from the manufacturing experts so that it can properly generate the decision to ‘scrap’, ‘rework’ or ‘use as is’ for the inspected parts. The real quality data from an oil field service industry is used to validate the effectiveness of the proposed tool.

Analysis of Professional Practices Inspectors of Education – Training in Morocco

IntroductionThis cross-sectional, descriptive and qualitative research had a threefold purpose: 1) to identify and describe different effective professional tasks and duties of the “expert” inspector of education; 2) contribute to the development of a competency framework of the inspector; 3) deduce the real training needs of inspectors. Method – Our study was conducted with two groups of people: the experienced inspectors (P1) and students - inspectors have completed initial training (P2). P1 was doubly solicited through a questionnaire ordinal Likert scale, and a semi structured interview. While P2 was solicited by a weekly board followed by a deferred interview in order to explain their pedagogical choices and their professional practices. The quantitative analysis of our data was performed by SPSS 20 software, while qualitative data were processed twice by the thematic content analysis (Bardin, 2007) and the cognitivo-discursive analysis (Ghiglione, 1998). Analysis of data collected by the weekly report was analyzed by the model of satellite information of the action (Vermersch, 2006). Results – In general, 62 professional tasks have been described by P1 against 51 tasks identified by P2. We identify four major groups of professional skills for the job of inspector among the P1; and five groups for the P2. According to P1, the tasks of control and inspection are predominant, followed by monitoring and evaluation tasks and those coaching-training. Whereas for P2, there is a predominance of missions successively methodological, institutional and relational character. At a second level, the inspector of education in Morocco performed globally and unevenly 10 roles. Conclusion – The daily practices of inspectors have largely evolved into tasks increasingly routine, administrative and relatively short. For such tasks, the inspector requires skills rather “technical”, disciplinary and incorporated. At the end of our study, we were able to identify ten major tasks of the inspector of education. The exploitation of all results allowed us to develop a professional profile of inspector of education in Morocco.

Defect inspection and extraction of the mobile phone cover glass based on the principal components analysis

Surface defect inspection is an important part of quality control in mobile phone cover glass manufacturing. The traditional method is usually carried out manually by experienced inspectors and thus lacks sufficient efficiency and accuracy. In this paper, an automatic defect inspection system based on the principal components analysis is proposed for five typical cover glass defects: scratch, crack, deformation, edge broken, and angle cutting. This inspection system is robust for the defect shapes and obtains high recognition accuracy. The inspection system includes three parts: pre-processing, PCA-based defect recognition, and defect edge extraction. After pre-processing, most of noise and outliers are eliminated and the pixels of defects in the image are well enhanced for inspection and recognition. The eigen-defect matrix is constructed to characterize the variation between defect images. Additionally, time consumption in constructing the eigen-defect matrix is also discussed. The experimental results show that the inspection system has achieved high accuracy for inspection and recognition.

A vision-aided automation system for destructive web thickness measurement of microdrills

Microdrills are fundamental cutting tools used in mechanical microdrilling. The web thickness of a microdrill is an important design parameter in balancing drill rigidity and chip removal ability. Conventional destructive method for measuring the web thickness is currently adopted by some microdrill manufactures. The conventional method must be carried out manually by experienced inspectors and thus lacks sufficient efficiency and accuracy. To improve the drawbacks of the conventional method, a vision-aided automation system for performing the destructive web thickness measurement is introduced in this paper. Based on the integration of motion and logic control, machine vision, and image processing techniques, an automated measuring process is then established. Experiments to measure the web thickness of microdrill samples were conducted to test the feasibility and effectiveness of the presented automation system and automated measuring process. The experimental results showed that the presented vision-aided automation system, in combination with the automated measuring process, could be an efficient and precise manner for replacing the conventional manual procedure in performing the destructive web thickness measurement of certain microdrills.

Effectiveness and Subjectivity of Visual Inspection as a Method to Assess Bending Stiffness and Strength of Chestnut Elements

Besides the difficulty of assessing an existing timber structure on site, the efficiency and accuracy of visual inspection is often compromised by its subjective nature inherited by the level of expertise of the inspector. This often leads to conservative predictions of the mechanical properties, even with the use of specific visual grading norms. The main objectives of this work are to assess the effectiveness of visual inspection as a method to define different classes of strength and stiffness and to provide a statistical analysis on its subjectivity. For that aim, visual inspection using Italian standard UNI 11119:2004 and bending tests of 20 old chestnut beams (Castanea sativa Mill.) at different scale element, were carried out. Comparisons and effectiveness of visual inspection is analyzed within and between different scales of the timber members, and also regarding different level of expertise of inspectors. The results evidence similar percentages of segments classified with higher and lower visual inspections classes and proved to be a good qualitative indicator of bending strength between sawn beams. An overall 42% accuracy of the most experienced inspectors was found with better differentiation between visual classes, whereas lower level inspectors scored approximately less 5%. Lower level inspectors also evidenced higher concentration of values around a higher mean for each class denoting a more conservative approach. Regardless of the inspector level, knot size was considered the main limiting visual parameter with higher influence in small scales of the timber elements. When studying the characterization of a single knot, coefficients of variation of 15.7% and 21.8% were found for measuring the minimum and maximum diameter. Bayesian probability networks were considered as to individually assess the accuracy in stiffness prediction of different level of inspectors, and by combination of their information, evidencing that parallel combination for prior information may allow the increase in visual inspection accuracy.