Search

With the advent of the Fourth Industrial Revolution, the economic, social, and technological demands for pipe maintenance are increasing due to the aging of the infrastructure caused by the increase in industrial development and the expansion of cities. Owing to this, an automatic pipe damage detection system was built using a laser-scanned pipe’s ultrasonic wave propagation imaging (UWPI) data and conventional neural network (CNN)-based object detection algorithms. The algorithm used in this study was EfficientDet-d0, a CNN-based object detection algorithm which uses the transfer learning method. As a result, the mean average precision (mAP) was measured to be 0.39. The result found was higher than COCO EfficientDet-d0 mAP, which is expected to enable the efficient maintenance of piping used in construction and many industries.

Given the explosive growth of information technology and the development of computer vision with convolutional neural networks, wildfire field data information systems are adopting automation and intelligence. However, some limitations remain in acquiring insights from data, such as the risk of overfitting caused by insufficient datasets. Moreover, most previous studies have only focused on detecting fires or smoke, whereas detecting persons and other objects of interest is equally crucial for wildfire response strategies. Therefore, this study developed a multilabel classification (MLC) model, which applies transfer learning and data augmentation and outputs multiple pieces of information on the same object or image. VGG-16, ResNet-50, and DenseNet-121 were used as pretrained models for transfer learning. The models were trained using the dataset constructed in this study and were compared based on various performance metrics. Moreover, the use of control variable methods revealed that transfer learning and data augmentation can perform better when used in the proposed MLC model. The resulting visualization is a heatmap processed from gradient-weighted class activation mapping that shows the reliability of predictions and the position of each class. The MLC model can address the limitations of existing forest fire identification algorithms, which mostly focuses on binary classification. This study can guide future research on implementing deep learning-based field image analysis and decision support systems in wildfire response work.

The atmospheric deposition of anthropogenic nitrogen is an increasingly important new source of nitrogen to the ocean. Coastal areas east of the Korean Peninsula are suitable for the investigation of the effects of atmospheric anthropogenic nitrogen on the ocean nutrient system because of the low riverine discharge rates and the prevailing influence of the East Asian outflow. Thus, we measured the concentrations of nitrate (NO3−) and ammonium (NH4+) in airborne particles and in precipitation from March 2014 to February 2016 at a coastal site (37.08°N, 129.41°E) on the east coast of Korea. The dry deposition of NO3− (27–30 mmol N m−2 yr−1) was far greater than that of NH4+ (6–8 mmol N m−2 yr−1). The greater rate of dry NO3− deposition was associated with air masses traveling over northeastern China and central Korea. In contrast, the rates of wet deposition of NO3− (17–24 mmol N m−2 yr−1) and NH4+ (14–27 mmol N m−2 yr−1) were comparable and were probably associated with in-cloud scavenging of these ions. The results indicate that the total deposition of NO3− and NH4+ combined could contribute to ~2.4% and ~1.9% of the primary production in the coastal areas east of the Korean Peninsula and in the East Asian marginal seas, respectively, which would be a lower bound because the dry deposition of reactive nitrogen gas was not included. Our study shows that the atmospheric input of anthropogenic NO3− and NH4+ may substantially increase phytoplankton biomass in the coastal waters of the East Sea near the Korean Peninsula.

It has been proposed that pre-stressed concrete bridges improve load performance by inducing axial pre-stress using pre-stress tendons. However, the tensile force of the pre-stress tendons could not be managed after construction, although it directly supports the load of the structure. Thus, the tensile force of the pre-stress tendon should be checked for structural health monitoring of pre-stressed concrete bridges. In this study, a machine learning–based tensile force estimation method for a pre-stressed concrete girder is proposed using an embedded elasto-magnetic sensor and machine learning method. The feedforward neural network and radial basis function network were applied to estimate the tensile force of the pre-stress tendon using the area ratio of the magnetic hysteresis curve measured by the embedded elasto-magnetic sensor. The feedforward neural network and radial basis function network were trained using 213 datasets obtained in laboratory experiments, and trained feedforward neural network and radial basis function network were applied to a 50-m real-scale pre-stressed concrete girder test for estimating tensile force. Nine embedded elasto-magnetic sensors were installed on the sheath, and the magnetic hysteresis curves of the pre-stress tendons were measured during tensioning. The area ratio was extracted and inputted to the trained feedforward neural network and radial basis function network to estimate the tensile force. The estimated tensile force was compared with the reference tensile force measured by the load cell. According to the result, the estimated tensile force can represent the actual tensile force of the pre-stress tendon without calibrating tensile force estimation algorithms at the site. In addition, it can measure the actual friction loss by estimating the tensile force at the maximum eccentric part. Based on the results, the proposed method might be a solution for the structural health monitoring of pre-stressed concrete bridges with field applicability.

This study investigates the applicability of an embedded EM sensor using a series of experimental studies. To verify the embedded EM sensor, the magnetic hysteresis of various types of PS tendons is measured. After that, the embedded EM sensor is embedded into the concrete and the possibility of obtaining measurements is verified. Finally, the downscaled PSC girder specimen having a sheath with a different curvature is fabricated and the influence of the sheath curvature is investigated. The magnetic hysteresis was changed constantly even though the type of PS tendon was changed, and the embedded EM sensor can measure the magnetic hysteresis, even in the concrete and curved sheath. The area of magnetic hysteresis was decreased according to the increase in the tension force, but the actual values were different according to the number and cross-sectional area of tendons and the initial state of sensors. To compensate for the measured data, the tensile force was converted to the tensile stress and the area ratio was used to compensate for the initial value of the EM sensor. According to the test results, the embedded EM sensor could be applied to the actual PS girder and it can measure the actual tension, which includes the friction loss.

Concrete strength and factors affecting its development during early concrete curing are important research topics. Avoiding uncertain incidents during construction and in service life of structures requires an appropriate monitoring system. Therefore, numerous techniques are used to monitor the health of a structure. This paper presents a nondestructive testing technique for monitoring the strength development of concrete at early curing ages. Dispersed carbon nanotubes (CNTs) were used with cementitious materials and piezoelectric (PZT) material, a PZT ceramic, owing to their properties of intra electromechanical effects and sensitivity to measure the electromechanical impedance (EMI) signatures and relevant properties related to concrete strength, such as the elastic modulus, displacement, acceleration, strength, and loading effects. Concrete compressive strength, hydration temperature, mixture ratio, and variation in data obtained from the impedance signatures using fuzzy logic were utilized in the comparative result prediction method for concrete strength. These results were calculated using a fuzzy logic-based model considering the maturity method, universal testing machine (UTM) data, and analyzed EMI data. In the study, for data acquisition, a hybrid PZT–CNT sensor and a temperature sensor (Smart Rock) were embedded in the concrete to obtain the hydration temperature history to utilize the concrete maturity method and provide data on the EMI signatures. The dynamic changes in the medium caused during the phase in the concrete strengthening process were analyzed to predict the strength development process of concrete at early curing ages. Because different parameters are considered while calculating the concrete strength, which is related to its mechanical properties, the proposed novel method considers that changes in the boundary condition occurring in the concrete paste modify the resonant frequency response of the structure. Thus, relating and analyzing this feature can help predict the concrete strength. A comprehensive comparison of the results calculated using the proposed module, maturity method, and cylindrical specimens tested using the UTM proved that it is a cost-effective and fast technique to estimate concrete strength to ensure a safe construction of reinforced cement concrete infrastructures.

We evaluated the potential impacts of atmospheric deposition on marine productivity and inorganic carbon chemistry in the northwestern Pacific Ocean (8-39°N, 125-157°E). The nutrient concentration in atmospheric total suspended particles decreased exponentially with increasing distance from the closest land-mass (Asia), clearly revealing anthropogenic and terrestrial contributions. The predicted mean depositional fluxes of inorganic nitrogen were approximately 34 and 15 μmol m-2 d-1 to the west and east of 140°E, respectively, which were at least two orders of magnitude greater than the inorganic phosphorus flux. On average, atmospheric particulate deposition would support 3-4% of the net primary production along the surveyed tracks, which is equivalent to ~2% of the dissolved carbon increment caused by the penetration of anthropogenic CO2. Our observations generally fell within the ranges observed over the past 18 years, despite an increasing trend of atmospheric pollution in the source regions during the same period, which implies high temporal and spatial variabilities of atmospheric nutrient concentration in the study area. Continued atmospheric anthropogenic nitrogen deposition may alter the relative abundances of nitrogen and phosphorus.

As civil engineering structures become larger, non-contact inspection technology is required to measure the overall shape and size of structures and evaluate safety. Structures are easily exposed to the external environment and may not be able to perform their original functions depending on the continuous load for a long time. Therefore, in this study, we propose a method for estimating the vertical displacement of structures using light detection and ranging, which enables non-contact measurement. The point cloud acquired through laser scanning was rearranged into a three-dimensional space, and internal nodes were created by continuously dividing the space. The generated node has its own location information, and the vertical displacement value was calculated by searching for the node where the deformation occurred. The performance of the proposed displacement estimation technique was verified through static loading experiments, and the octree space partitioning method is expected to be applied and utilized in structural health monitoring.

Climate change increases the frequency and intensity of heatwaves, causing significant human and material losses every year. Big data, whose volumes are rapidly increasing, are expected to be used for preemptive responses. However, human cognitive abilities are limited, which can lead to ineffective decision making during disaster responses when artificial intelligence-based analysis models are not employed. Existing prediction models have limitations with regard to their validation, and most models focus only on heat-associated deaths. In this study, a random forest model was developed for the weekly prediction of heat-related damages on the basis of four years (2015&ndash, 2018) of statistical, meteorological, and floating population data from South Korea. The model was evaluated through comparisons with other traditional regression models in terms of mean absolute error, root mean squared error, root mean squared logarithmic error, and coefficient of determination (R2). In a comparative analysis with observed values, the proposed model showed an R2 value of 0.804. The results show that the proposed model outperforms existing models. They also show that the floating population variable collected from mobile global positioning systems contributes more to predictions than the aggregate population variable.

In the construction industry, it is difficult to predict occupational accidents because various accident characteristics arise simultaneously and organically in different types of work. Furthermore, even when analyzing occupational accident data, it is difficult to deduce meaningful results because the data recorded by the incident investigator are qualitative and include a wide variety of data types and categories. Recently, numerous studies have used machine learning to analyze the correlations in such complex construction accident data, however, heretofore the focus has been on predicting severity with various variables, and several limitations remain when deriving the correlations between features from various variables. Thus, this paper proposes a data processing procedure that can efficiently manipulate accident data using optimal machine learning techniques and derive and systematize meaningful variables to rationally approach such complex problems. In particular, among the various variables, the most influential variables are derived through methods such as clustering, chi-square, Cramer&rsquo, s V, and predictor importance, then, the analysis is simplified by optimally grouping the variables. For accident data with optimal variables and elements, a predictive model is constructed between variables, using a support vector machine and decision-tree-based ensemble, then, the correlation between the dependent and independent variables is analyzed through an alluvial flow diagram for several cases. Therefore, a new processing procedure has been introduced in data preprocessing and accident prediction modelling to overcome difficulties from complex and diverse construction occupational accident data, and effective accident prevention is possible by deriving correlations of construction accidents using this process.

To better understand carbon cycles in coastal and marginal seas, time-series monitoring is essential because of large temporal variabilities. In this regard, we conducted monthly field researches from April 2017 to May 2019 at the Socheongcho (SCC) Ocean Research Site (37°N‚124°E) in the Yellow Sea located between Korea and China. At each survey, we collected surface seawater samples during approximately 7 days with an sampling interval of two hours (except for spring 2017). Total alkalinity (TA) and dissolved inorganic carbon (DIC) were analyzed by using VINDTA 3C system, Apollo SciTech DIC analyzer and Apollo SciTech Alkalinity Titrator. In addition, a pH sensor (SeapHOx) was installed at the surface layer from September 2018 to June 2019 which is also capable of measuring salinity, temperature and oxygen. Based on the observations, we estimated a partial pressure of carbon dioxide (pCO2) and aragonite saturation state. As expected, seasonal variations in TA and DIC were strongly associated with those of salinity. We also detected a sudden increase DIC in October when vertical mixing was greatly enhanced. Despite a large outgassing during the fall season, annual mean air--sea influx of CO2 was ∼0.61mol·m−2·year−1, suggesting that the study area was a weak sink for atmospheric CO2. Aragonite was enerally reduced during winter (∼1.5). However, no undersaturation event was found during the whole investigation.

Corrosion detection using a pulsed laser scanning system can be performed via ultrasonic wave propagation imaging. This method outputs illustrations of the wave field within the host structure, thus, it can depict wave&ndash, corrosion area interactions. Additionally, post-processing can be performed to enhance the visualization of corroded areas. The wavefield energy computed using RMS (Root Mean Square) is a validated post-processing tool capable of displaying the location and area of corrosion-damaged regions. Nonetheless, to characterize corrosion, it is necessary to determine its depth. The measurement of depth in conjunction with that of the corroded area via the RMS distribution enables the determination of all dimensions of corrosion damage. Thereafter, the flaw severity can be evaluated. This study employed a wavefield within a plate on which corrosion was developed artificially to generate frequency&ndash, wavenumber dispersion curves. The curves were compared with their counterparts from a corrosion-free plate. Alternatively, they could be compared with dispersion curves drawn using the depth and material properties of a pristine plate via a computer program. Frequency&ndash, wavenumber pairs were extracted from the dispersion curves produced using the portion of the wavefield within the corroded area. These were inserted into the Rayleigh&ndash, Lamb equation, from which depths were calculated and averaged.

A terrestrial light detection and ranging (LiDAR) can be used to construct the building information modeling as well as to measure shape deformation that varies with time or load during or ...

In this study, the hardening process of ultra high performance concrete (UHPC) was monitored non-destructively using a single embedded sensor system and the characteristics of guided waves, especially the Lamb wave. Lamb wave propagation depends on the material properties of the medium and boundary conditions. Since the boundary conditions of the embedded sensor system continuously change during the hardening process of concrete materials, the measured characteristics of the propagating waves also vary. To understand the variations in wave propagation, the Lamb modes were decomposed using the polarization characteristics of piezoelectric sensors, which were used to measure wave responses. Additionally, a traditional penetration resistance method was adopted to estimate the time for phase transition of UHPC. The decomposed Lamb modes were compared to measurements of penetration resistance. The strength development of UHPC, with and without short-fiber reinforcement, was estimated using the variation of patterns of the decomposed Lamb modes after the phase transition. Based on the proposed methodology, which measures the propagation and variation of the Lamb waves, it is possible to estimate the time of phase transition and the strength development of UHPC.

It has been found that valence and arousal are the core affect dimensions in emotional structure. In this study, we hypothesized that there might be differences between different age groups in emotional structure using six facial expression stimuli (angry, disgusted, fearful, neutral, happy, and sad) of three age groups (young, middle-aged, and old). Unlike previous studies asking participants to rate subjective ratings or similarities between stimuli, participants in this study were required to determine whether stimulus pairs were the same or different emotions and reaction time and accuracy were measured for further analyses. We assumed that it would be harder when the stimulus pair is similar whereas it would be easier when the pair is different. The results showed that for the same emotion pair condition, the sad-sad pair had the lowest accuracy and the longest reaction time, while the happiness-happiness pair had the highest accuracy and the shortest reaction time. For the different emotion pairs, angry-disgusted and disgusted-sad was the lowest accuracy and the longest reaction time. For age of the stimuli effect, responses to the old faces had the lowest accuracy and the longest reaction time. The results suggest that identification of emotional stimuli might be affected by emotion category and age. Further study may need to recruit various age groups, because participants in the current study were mostly young adults.

본 연구에서는 누설자속기법을 이용한 와이어 로프 압착 변형 진단을 하였다. 와이어 로프를 사용하는 구조물에 적용하기 위하여 리프트오프의 발생을 최소화시킨 4채널의 누설자속센서 헤드를 제작하였다. 이를 사용하여 와이어 로프의 압착 변형 검색 실험을 수행하였다. 압착 변형 검색 실험을 위해 80 cm의 와이어 로프를 준비하였고, 중간인 40 cm 지점에 압착 변형을 가하였다. 와이어 로프 압착 변형을 진단하기 위해 자속신호를 힐버트 변환을 활용하여 순간변화량을 손상 검색에 활용하였다. 객관적인 손상 판단을 위하여 계측된 자속신호를 통계학적 GEV(generalized extreme value) 분포를 이용하여 임계값을 설정하였다. 이에 압착 손상을 가한 와이어 로프의 양수화된 데이터에 임계값을 적용해본 결과 압착 변형의 손상 유무를 확인할 수 있었다. 최종적으로 신호의 크기가 압착 변형의 돌출부보다 압축부에서 더 크게 나타난 것을 확인하였으며, 본 기법의 압착 변형 검색 가능성을 살펴보았다.

A magnetic flux leakage method was applied to detect damage when inspecting steel wire rope. A multi-channel magnetic flux leakage sensor head was fabricated using Hall sensors and permanent magnets to adapt to the wire rope. Three types of artificial damage were created on a wire rope specimen. The magnetic flux leakage sensor head scanned the damaged specimen to measure the magnetic flux density while the damage was expanding in three steps. Signal processing processes including the enveloping process based on Hilbert transform were performed to clarify the flux leakage signal due to the damage. The enveloped signals were then analyzed for objective damage detection by comparing with the threshold value. For improvement of quantitative analysis, three types of new damage indexes that utilize the relationship between the enveloped magnetic flux leakage signal and the threshold value were additionally proposed. By using the proposed damage indexes and the general damage indexes for the magnetic flux leakage method, the detected magnetic flux leakage signals from each damage type were quantified. The trends of the extracted damage indexes according to damage levels were analyzed to examine the applicability and reliability of the proposed damage indexes for the magnetic flux leakage based wire rope inspection.

이러한 연속체 구조물의 경우 대부분 외부환경에 노출되어 있어 다양한 환경적, 물리적 요인으로 인해 부식, 크랙, 단선과 같은 국부손상이 발생하게 될 수 있는데 이러한 손상에 대한 초기 대처가 늦을 경우 특정부의 응력 집중, 유체누출 등과 같은 원인으로 구조물 전체의 대형 사고로 이어질 수 있는 위험성을 가지고 있다.따라서 이러한 위험성을 조기에 방지하기 위해 부재에 발생한 손상을 초기에 발견할 수 있는 검사법이 필수적인데, 짧게는 수십 미터에서 길게는 수천 킬로미터까지 이어지는 거대한 연속체 구조물의 특성상, 적용되는 검사법은 신뢰성과 더불어 경제성을 반드시 요구하게 된다. 이에 뛰어난 신뢰도와 높은 재현성의 장점을 지니고 있는 마그네틱 센서 응용기술들이 항공, 선박, 교량 케이블 등을 포함한 다양한 강자성체 구조물의 비파괴 평가에 매우 널리 활용되어 왔다[1-4]. 마그네틱 센서는 그 종류가 다양하므로 계측하고자 하는 대상 구조물의 구조적 특성에 맞춰 사용하도록 되어 있는데, 그 중 파이프 및 와이어와 같은 연속체 구조물을 대상으로 한 검사에 있어서는 자속누설탐상법을 이용한 검사가 주를 이루고 있다[5-9].자속누설탐상법은 포신의 결함 탐상을 위해 1868년 영국에서 처음으로 개발되었는데 포신을 자화시켜 나침반을 이용하여 표면을 검사하는데 사용되었었다[10]. 20세기에 자기센서가 처음으로 개발된 이후, 자속누설탐상법은 현재까지 지속적으로 많은 발전을 이루었으며 그 적용 범위 역시 매우 넓은 영역으로 확대되고 있다[11].1. 자속누설탐상법의 기본 원리강재 시편이 자화되어 있을 때 강재는 자석과 같은 특성을 띄게 되는데, 이러한 강재에 국부손상이 발생했을 때 결함 부분에서 자기장의 통과 단면적이 줄어들게 되고 손상 틈의 공기는 자석만큼의 자기장 밀도를 가지지 못하기 때문에 그림 2와 같이 손상부에 자속누설이 발생하게 된다. 이러한 원리에 기초하여 자속의 누설을 비접촉 방식으로 계측함으로써 강재의 국부손상을 검색할 수 있다. 자속누설의 비접촉 계측 원리는 그림 3에서 보는 바와 같이 영구자석에 의하여

Recently, studies on cementless concrete with the advantages of greenhouse gas reduction and high strength have been actively pursued as the public interest in the development of environmentally friendly concrete and increase in the life expectancy of the structure has increased. On the other hand, in the case of cementless concrete, it is essential to observe the strength change at the initial stage of curing and to measure the precise strength because there is a risk of damage and collapse of the structure due to the sudden drop in strength of the concrete during or after construction. In this study, wave mode decomposition method was used to estimate the early-age strength of alkali activated concrete. Experiments were carried out to observe that the shape of the wave measured by each sensor changes as the internal temperature of the concrete rises due to internal hydration heat following the insertion of the embedded piezoelectric sensor module into the specimen immediately after the concrete pouring. As the time elapsed, it was observed that the arrival time and amplitude of the wave were changing. So it was confirmed that the early-age strength of concrete can be estimated through this study.

In recent scenario, the information about buried objects needed for redevelopment and reorganization of the complicated urban environment. Accidents caused by pipeline damages such as gas, communication and underground electric power lines result in human and financial loss. Therefore, the information of underground obscured materials is essential for smart city realization and construction. GPR (Ground Penetrating Radar) investigation has advantages of high resolution detection, ease of utilization and strong electromagnetic signal to noise ratio when using high frequency. However, the GPR detected image data is not visible and has a problem that it may be interpreted differently based on the approaching skill of the inspector. Therefore, this study was conducted to verify the visualization of detected data using computer vision based on the data from GPR. Canny edge and Harris corner detection were applied to the GPR image data to detect the hyperbolic shape. Using these techniques to enhance visibility will contribute to the reliable result in the buried materials detection.

As high-rise buildings and aging structures increase, researches are underway to effectively inspect structures and prevent accidents in Korea. The structure cannot maintain its original shape due to natural disasters and loads, and deformation occurs. The accidents caused by the deformation of the structure have become a common news in the surroundings. Therefore, this study proposes a noncontact shape management monitoring method that can check structures and detect deformation.

To detect local damage for inspecting the steel chains of the escalator, a magnetic flux leakage (MFL) method was used in this study. To demonstrate the proposed damage detection method through experiments, a multi-channel MFL sensor head was fabricated using a Hall sensor array and magnetic yokes to adapt to the steel chain. To prepare the damaged steel chain specimens, several amounts of artificial damages were formed on steel chains. The MFL sensor head scanned the damaged specimens to measure the magnetic flux signals. After obtaining the signals, a series of signal processing processes, including the enveloping process based on the Hilbert transform (HT), was performed to better recognize the MFL signals by reducing the unexpected noise. The enveloped signals were then analysed for objective damage detection by comparing them with a threshold that was established based on the generalized extreme value (GEV) distribution. The detected MFL signals that exceed the threshold were analysed quantitatively by extracting the magnetic features from the MFL signals. For the quantitative analysis, damage indexes based on the enveloped MFL signal and the threshold value were extracted. By using the damage indexes for the MFL method, the detected MFL signals for each damage type were quantified. Finally, multi-stage pattern recognition method using extracted multi-scale damage indexes was implemented to automatically estimate the information of damage. To analyse the reliability of the MFL based automated steel chain NDE method, the accuracy and reliability were evaluated by comparing the repeatedly estimated damage information and the actual damage.

The actual tensile force of pre-stressed (PS) tendons of a pre-stressed concrete (PSC) girder is one of the important factors for evaluating the performance of PSC girder bridges. To measure the tensile force of the PS tendon, this study proposed a radial basis function network (RBFN) based tensile force estimation method using embedded elasto-magnetic (EM) sensors. The magnetic hysteresis of PS tendons are changed according to the applied tensile force. To measure the magnetic hysteresis of PS tendon of PSC girder, the EM sensor should be embedded in the PSC girder because the PS tendons were located in inside of PSC girder. The RBFN was used to estimate the tensile force using the variations in magnetic hysteresis. To verify the proposed method, the in-field tests were performed. The embedded EM sensors were embedded into PSC girder specimen and the magnetic hysteresis changes due to the variations in tensile forces were measured using embedded EM sensors. The tensile forces were estimated using trained RBFN and they compared with reference tensile forces measured by hydraulic jacking machine. According to the measurement results, the proposed method can be a one of the solution to monitor the tensile force of PS tendons.

In the real world, the main cables of suspension bridges are commonly inspected by conducting a periodic visual inspection of the exterior cover of the cable. Although there is a need to conduct a nondestructive evaluation (NDE) of the damage of the main cable, a suitable NDE technique has not yet been developed due to the large diameter and low accessibility of the cable. This study investigates a magnetic sensing cross-sectional loss quantification method that can detect internal and external damage to the main cables. This main cable NDE method applies an extremely low-frequency alternating current (ELF-AC) magnetization method and search coil sensor-based total flux measurement. A total flux sensor head consists of a magnetization yoke and a search coil sensor. To magnetize the main cable, a magnetic field was generated by applying a triangular ELF-AC voltage to the electromagnet yoke. The sensing part measures the magnetic flux that passes through the search coil, and the B-H loop was then obtained using the relationship between the ELF-AC voltage that has been input and the total flux that was measured. Also, the cross-sectional loss can be quantified using a variation of magnetic features from the B-H loop. To verify the feasibility of using the proposed NDE technique, a series of experiments were performed using a main cable specimen with a gradual increase in the cross-sectional loss. Finally, the relationship between the cross-sectional loss and extracted magnetic feature was determined and used to quantify the cross-sectional loss via the proposed method.

As the frequency of earthquakes has increased in Korea in recent years, designing earthquake-resistant facilities has been increasingly emphasized. Structures constructed with rebars are vulnerable to shaking, which reduces their seismic performance and may result in damage to human life and property. Because the construction of facilities requires the maintenance of sub-constructions, such as by cutting rebars or compensating for missing rebars, information on rebar diameter is required. In this study, the YOLO-v3 algorithm, which has the fastest object recognition performance, was applied to the structural correction data, and a basic experiment was conducted in the air to predict the diameter of rebars in a facility, in real time based on ground-penetrating radar data. The reason for using the YOLO-v3 algorithm is that in the case of GPR data that change slightly according to the diameter of the reinforcing bar, it is difficult to discriminate with the naked eye, and the result may change depending on the inspector. The model achieved a higher accuracy than conventional rebar detection and diameter prediction methods. In addition, the possibility of real-time rebar diameter prediction during construction, using the proposed method, was verified.

Modeling a structure in the virtual world using three-dimensional (3D) information enhances our understanding, while also aiding in the visualization, of how a structure reacts to any disturbance. Generally, 3D point clouds are used for determining structural behavioral changes. Light detection and ranging (LiDAR) is one of the crucial ways by which a 3D point cloud dataset can be generated. Additionally, 3D cameras are commonly used to develop a point cloud containing many points on the external surface of an object around it. The main objective of this study was to compare the performance of optical sensors, namely a depth camera (DC) and terrestrial laser scanner (TLS) in estimating structural deflection. We also utilized bilateral filtering techniques, which are commonly used in image processing, on the point cloud data for enhancing their accuracy and increasing the application prospects of these sensors in structure health monitoring. The results from these sensors were validated by comparing them with the outputs from a linear variable differential transformer sensor, which was mounted on the beam during an indoor experiment. The results showed that the datasets obtained from both the sensors were acceptable for nominal deflections of 3 mm and above because the error range was less than &plusmn, 10%. However, the result obtained from the TLS were better than those obtained from the DC.

Estimating the damaged area after a forest fire is important for responding to this natural catastrophe. With the support of aerial remote sensing, typically with unmanned aerial vehicles (UAVs), the aerial imagery of forest-fire areas can be easily obtained; however, retrieving the burnt area from the image is still a challenge. We implemented a new approach for segmenting burnt areas from UAV images using deep learning algorithms. First, the data were collected from a forest fire in Andong, the Republic of Korea, in April 2020. Then, the proposed two-patch-level deep-learning models were implemented. A patch-level 1 network was trained using the UNet++ architecture. The output prediction of this network was used as a position input for the second network, which used UNet. It took the reference position from the first network as its input and refined the results. Finally, the final performance of our proposed method was compared with a state-of-the-art image-segmentation algorithm to prove its robustness. Comparative research on the loss functions was also performed. Our proposed approach demonstrated its effectiveness in extracting burnt areas from UAV images and can contribute to estimating maps showing the areas damaged by forest fires.

In recent years, building energy consumption has increased in response to climate change, resulting in a worldwide reduction of energy efficiency. A strong response is required, through both the enhancement of building energy efficiency and the reduction of power usage. These objectives could be achieved by pursuing “Blackout” prevention, through developing a control unit such as that of the urban energy grid system that was used in the Energy Operation Center (EOC) for smart cities. In this context, this paper presents an optimization operation technique for Building Energy Management Systems (BEMS), using control monitor-based Building Information Modeling (BIM) for the efficient operation of the EOC. BIM is one of the approaches that can be used for the visual representation, management, and exchange of information on all aspects of a building. The EOC should be able to efficiently represent the data from the building information, and the operator maintaining it should be able to promptly acquire the data for handling it. This suggests that the control of energy consumption by a Building Automation System (BAS) to maximize building energy efficiency will lead to improved total energy performance, reduced operating costs, and reduced environmental impact.

This paper describes a smart structural system, which uses smart materials for real-time monitoring and active control of bolted-joints in steel structures. The goal of this research is to reduce the possibility of failure and the cost of maintenance of steel structures such as bridges, electricity pylons, steel lattice towers and so on. The concept of the smart structural system combines impedance based health monitoring techniques with a shape memory alloy (SMA) washer to restore the tension of the loosened bolt. The impedance-based structural health monitoring (SHM) techniques were used to detect loosened bolts in bolted-joints. By comparing electrical impedance signatures measured from a potentially damage structure with baseline data obtained from the pristine structure, the bolt loosening damage could be detected. An outlier analysis, using generalized extreme value (GEV) distribution, providing optimal decision boundaries, has been carried out for more systematic damage detection. Once the loosening damage was detected in the bolted joint, the external heater, which was bonded to the SMA washer, actuated the washer. Then, the heated SMA washer expanded axially and adjusted the bolt tension to restore the lost torque. Additionally, temperature variation due to the heater was compensated by applying the effective frequency shift (EFS) algorithm to improve the performance of the diagnostic results. An experimental study was conducted by integrating the piezoelectric material based structural health monitoring and the SMA-based active control function on a bolted joint, after which the performance of the smart \'self-monitoring and self-healing bolted joint system\' was demonstrated.

Generally, many previous researches conducted using FBG sensors were based on the assumption that the coefficient of thermal expansion (CTE) and the thermo-optic coefficient (TOC) of FBG sensors are constants. However, they should be considered variables for higher accuracy when FBG sensors are applied to circumstances with temperature changes. In this paper, CTE and TOC of FBG sensors were precisely investigated from ambient room temperature (293 K) to cryogenic temperature (113 K). For this purpose, a new non-linear relationship between the wavelength shift and temperature change was proposed. From the test results, it was successfully investigated that the sum of CTE and TOC had a non-linear dependency on temperature changes and varied from 5.9×10-6 /K (293 K) to 1.6×10-6 /K (113 K).

In this paper, an embedded EM sensor was researched to estimate prestressing force of PS tendon in PSC girder. Recent methodologies for managing prestressing force loss were staying on verifying a applying prestressing force under construction, namely the loss management can not be controlled after construction. To estimate the tensile force of PS tendon during lifetime of PSC girder, this research proposed a bobbin-type embedded EM sensor that can be embedded in PSC girder is designed and fabricated considering the shape properties of anchorage zone and sheath. To verify the proposed sensor, a small PSC girder test was performed. The embedded EM sensor was connected to a sheath and anchor block, and the concrete was poured. After curing, the change of the permeability of PS tendon under tensile forces of 200, 710, 1070, 1300kN was measured using embedded EM sensor. The permeability of PS tendon had decreased according to the increment of applied tensile force. Also it is confirmed that the change of permeability due to applied tensile force could resolve the applied tensile force values. As a result, proposed embedded EM sensor could be embed into the PSC girder and it could be used to estimate the tensile force variation during lifetime of PSC girder.

To minimize the damage caused by wildfires, a deep learning-based wildfire-detection technology that extracts features and patterns from surveillance camera images was developed. However, many studies related to wildfire-image classification based on deep learning have highlighted the problem of data imbalance between wildfire-image data and forest-image data. This data imbalance causes model performance degradation. In this study, wildfire images were generated using a cycle-consistent generative adversarial network (CycleGAN) to eliminate data imbalances. In addition, a densely-connected-convolutional-networks-based (DenseNet-based) framework was proposed and its performance was compared with pre-trained models. While training with a train set containing an image generated by a GAN in the proposed DenseNet-based model, the best performance result value was realized among the models with an accuracy of 98.27% and an F1 score of 98.16, obtained using the test dataset. Finally, this trained model was applied to high-quality drone images of wildfires. The experimental results showed that the proposed framework demonstrated high wildfire-detection accuracy.

The application of deep learning (DL) algorithms to non-destructive evaluation (NDE) is now becoming one of the most attractive topics in this field. As a contribution to such research, this study aims to investigate the application of DL algorithms for detecting and estimating the looseness in bolted joints using a laser ultrasonic technique. This research was conducted based on a hypothesis regarding the relationship between the true contact area of the bolt head-plate and the guided wave energy lost while the ultrasonic waves pass through it. First, a Q-switched Nd:YAG pulsed laser and an acoustic emission sensor were used as exciting and sensing ultrasonic signals, respectively. Then, a 3D full-field ultrasonic data set was created using an ultrasonic wave propagation imaging (UWPI) process, after which several signal processing techniques were applied to generate the processed data. By using a deep convolutional neural network (DCNN) with a VGG-like architecture based regression model, the estimated error was calculated to compare the performance of a DCNN on different processed data set. The proposed approach was also compared with a K-nearest neighbor, support vector regression, and deep artificial neural network for regression to demonstrate its robustness. Consequently, it was found that the proposed approach shows potential for the incorporation of laser-generated ultrasound and DL algorithms. In addition, the signal processing technique has been shown to have an important impact on the DL performance for automatic looseness estimation.