• Journal of the Society of Disaster Information
• /
• v.19 no.3
• /
• pp.498-509
• /
• 2023

Purpose: This paper proposes a method detecting the falling of a maintenance worker in the underground utility tunnel, by applying deep learning techniques using CCTV video, and evaluates the applicability of the proposed method to the worker monitoring of the utility tunnel. Method: Each rule was designed to detect the falling of a maintenance worker by using the inference results from pre-trained YOLOv5 and OpenPose models, respectively. The rules were then integrally applied to detect worker falls within the utility tunnel. Result: Although the worker presence and falling were detected by the proposed model, the inference results were dependent on both the distance between the worker and CCTV and the falling direction of the worker. Additionally, the falling detection system using YOLOv5 shows superior performance, due to its lower dependence on distance and fall direction, compared to the OpenPose-based. Consequently, results from the fall detection using the integrated dual deep learning model were dependent on the YOLOv5 detection performance. Conclusion: The proposed hybrid model shows detecting an abnormal worker in the utility tunnel but the improvement of the model was meaningless compared to the single model based YOLOv5 due to severe differences in detection performance between each deep learning model

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.19 no.1
• /
• pp.1-9
• /
• 2017

In this study, a test-bed system simulated a tunnel and concrete track is tested on cite and invested an allowed limit of multi-layered sound absorbing panel for reducing noise reflected on the concrete track in train tunnel considering the criteria and limitation on the theoretical back ground. The studied results are an effective evaluating system of the sound absorbing coefficient influenced fluid effects depending on the vehicle speed in the urban train tunnel and measuring not only structural behaviors of maximum displacement and acceleration of the panel but also dynamic characteristics of damping ratio and natural frequency.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2009.09a
• /
• pp.117-129
• /
• 2009

This paper studied the safety management network system of infrastructure which constructed smart sensors, closed-circuit television(CCTV) and monitoring system. This safety management of infrastructure applied to bridge, cut slop and tunnel, embankment etc. The system applied to technologies of standardization guidelines, data acquirement technologies, data analysis and judgment technologies, system integration setup technology, and IT technologies. It was constructed safety management network system of various infrastructure to improve efficient management and operation for many infrastructure. Integrated safety management network system of infrastructure consisted of the real-time structural health monitoring system of each infrastructure, integrated control center, measured data transmission using i of tet web-based, collecting data using sf ver, early alarm system which the dangerous event of infrastructure occurred. Integrated control center consisted of conference room, control room to manage and analysis the data, server room to present the measured data and to collect the raw data. Early alarm system proposed realization of warning and response within 5 minute or less through development of sensor-based progress report and propagation automation system using the media such as MMS, VMS, EMS, FMS, SMS and web services of report and propagation. Based on this, the most effective u-Infrastructure Safety Management System is expected to be stably established at a less cost, thus making people's life more comfortable. Information obtained from such systems could be useful for maintenance or structural safety evaluation of existing structures, rapid evaluation of conditions of damaged structures after an earthquake, estimation of residual life of structures, repair and retrofitting of structures, maintenance, management or rehabilitation of historical structures.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.10 no.4
• /
• pp.160-166
• /
• 2006

Optical Fiber Bragg Grating sensor has a good performance to measure microscopic displacement which can measure strain of lining concrete and cylindrical structure like high intensity containment building and it can present many advantages like a corrosion resistance from the durability point of view. Then it can measure plane geometrical displacement of cylindrical structures with two-way displacement FBG sensor module. According to the test result, measurement of FBG sensor is better performance than other electric sensor system and 2D-level measurement. As a test result, Resolution of the two-way displacement sensor module with FBG sensors are more 10 times than other LVDT or 2D surveying.

• Tunnel and Underground Space
• /
• v.8 no.1
• /
• pp.37-45
• /
• 1998

In this study, two dimensional and three dimensional numerical analysis were performed with a finite difference code for the safe maintenance of the spillway slope of the Boryeong dam. Results of the geological survey and the stereographic projection analysis on the discontinuities were used as input data for the numerical analysis. As a result, several suggestions were given such as the reinforcement of the local tension zone, the decrease in the angle of the slope, the drawdown of the pore pressure in the slope and the removal of the upper benches. A systematic and long-term monitoring system was also suggested.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.18 no.5
• /
• pp.128-134
• /
• 2014

Recently, design and construction of street tunnels tend to focus on cost reduction and preservation of nature. Accordingly, research is actively being carried out to quickly detect fires when they occur in tunnels, which have partially closed structures. Among such research, fire detection methods using optical fiber sensors have a wide bandwidth and fast transmission speed, while using light as a medium. Therefore, it does not receive electrical interference and there is almost no loss of information during transmission, while also having little noise as well. In relation to this, a fire monitoring system that can accurately detect the location of fires in real time using shape memory alloy and optical cables was developed in this study. In order to verify the developed method, light loss measurement test was conducted according to indoor temperature changes, while also conducting fire simulation tests by installing test beds in common underground zones with different external environments of temperature and distance. Upon carrying out experiments, the fire monitoring system developed in this study was found to be able to detect fires in long distance sections in real time.

• Tunnel and Underground Space
• /
• v.20 no.3
• /
• pp.207-216
• /
• 2010

Rockfall monitoring systems generally used in the country are mainly based on the detection of tension of protection wire or tilting of protection post due to rock fall. However, rock fall protection net must be installed prior to the monitoring system and continual maintenance work after each rock fall event is required for a normal operation of these detection systems. To solve these problems, we suggested and implemented a non-contact rock fall detection system using multiple photo sensors and additional camera. After a laboratory experiment and field application, we can conclude that this system is effective and reliable for detecting, collecting and analyzing the rock fall information. In addition, lighten and difference operations on two captured images were able to yield rough estimation of size and direction of rock fall.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2008.03a
• /
• pp.1278-1289
• /
• 2008

Currently in increasing number of urban tunnels with small overburden are excavated according to the principle of the New Austrian Tunneling Method (NATM). Successful design, construction and maintenance of NATM tunnel in urban area demands prediction, control and monitoring of surface settlement, gradient and ground displacement with high accuracy. Use of measured displacement for parameter determination has been researched over the years, and one geotechnical engineering principle has been formed as back analysis. In this paper, back analysis of a ground deformational behavior involving nonlinear behavior is discussed. It is of primary importance to make reliable prediction of deformational behavior for shallow tunnels in soft ground. However, predictions made often prove to be incorrect due to complexity of constitutive law and other relevant factors. Back analysis therefore becomes more important, for it may be used to interpret measured displacement to derive nonlinear material characteristics. The paper shows some example in which a deformational mechanism is studied in the light of inhomogeneous distrubution of Young's module, from which a logic is derived to identify two different types of nonlinear constitutive relationships.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.8
• /
• pp.8-16
• /
• 2016

This study presents the visualization of shape information based on Octree using 3D laser scanning. The process of visualization was established to construct the Octree structure from the 3D scan data. The scan data was converted to a 2D surface through the mesh technique and the surface was then converted to a 3D object through the Raster/Vector transformation. The 3D object was transmitted to the Octree Root Node and The shape information was constructed by the recursive partitioning of the Octree Root Node. The test-bed was selected as the steel bridge structure in Sungkyunkwan University. The shape information based on Octree was condensed into 89.3%. In addition, the Octree compressibility was confirmed to compare the shape information of the office building, a computer science campus in Germany and a New College in USA. The basis is created by the visualization of shape information for double-deck tunnel and it will be expected to improve the efficiency of structural health monitoring and maintenance.

• Wind and Structures
• /
• v.36 no.4
• /
• pp.237-247
• /
• 2023

The emergence of high-rise buildings has necessitated frequent structural health monitoring and maintenance for safety reasons. Wind causes damage and structural changes on tall structures; thus, safe structures should be designed. The pressure developed on tall buildings has been utilized in previous research studies to assess the impacts of wind on structures. The wind tunnel test is a primary research method commonly used to quantify the aerodynamic characteristics of high-rise buildings. Wind pressure is measured by placing pressure sensor taps at different locations on tall buildings, and the collected data are used for analysis. However, sensors may malfunction and produce erroneous data; these data losses make it difficult to analyze aerodynamic properties. Therefore, it is essential to generate missing data relative to the original data obtained from neighboring pressure sensor taps at various intervals. This study proposes a deep learning-based, deep convolutional generative adversarial network (DCGAN) to restore missing data associated with faulty pressure sensors installed on high-rise buildings. The performance of the proposed DCGAN is validated by using a standard imputation model known as the generative adversarial imputation network (GAIN). The average mean-square error (AMSE) and average R-squared (ARSE) are used as performance metrics. The calculated ARSE values by DCGAN on the building model's front, backside, left, and right sides are 0.970, 0.972, 0.984 and 0.978, respectively. The AMSE produced by DCGAN on four sides of the building model is 0.008, 0.010, 0.015 and 0.014. The average standard deviation of the actual measures of the pressure sensors on four sides of the model were 0.1738, 0.1758, 0.2234 and 0.2278. The average standard deviation of the pressure values generated by the proposed DCGAN imputation model was closer to that of the measured actual with values of 0.1736,0.1746,0.2191, and 0.2239 on four sides, respectively. In comparison, the standard deviation of the values predicted by GAIN are 0.1726,0.1735,0.2161, and 0.2209, which is far from actual values. The results demonstrate that DCGAN model fits better for data imputation than the GAIN model with improved accuracy and fewer error rates. Additionally, the DCGAN is utilized to estimate the wind pressure in regions of buildings where no pressure sensor taps are available; the model yielded greater prediction accuracy than GAIN.