Intelligence, Informatics and Infrastructure
Online ISSN : 2758-5816
Current issue
Displaying 1-11 of 11 articles from this issue
  • Monami Aoyama, Atsushi Hashimoto, Mayuko Yagishita
    2024 Volume 5 Issue 2 Pages 1-9
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    The interpretation of measurement result data is crucial in non-target analysis, an approach that has gained prominence in recent years for screening chemical substances in environmental samples. In this context, the authors have proposed a novel method that utilizes machine learning and image classification to analyze excitation-emission matrix (EEM) spectrum image data, offering a streamlined approach for screening environmental samples. This study specifically explored the viability of using AI to identify EEM spectrum image data from river water samples, both with and without added pesticides. Additionally, the qualitative and quantitative efficacy of image data as training data was scrutinized. The findings indicated that this method could be employed as a straightforward screening technique. However, merely increasing the volume of data derived from precise EEM spectrum measurements does not automatically enhance the accuracy of AI-based decisions. This highlights a critical aspect of data analysis in non-target screening methods, highlighting the importance of not only data quantity but also its relevance and quality in improving AI-driven analytical processes.

  • Riku Miyakawa, Takumi Murai, Yoshito Saito, Kenta Itakura
    2024 Volume 5 Issue 2 Pages 10-21
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    Accurate soybean sorting is an important but time- and labor-intensive process in soybean production. Therefore, inexpensive and accurate sorting machines are needed. Many of the currently reported models for discriminating external defects in soybeans have used deep learning, but the high cost of deep learning is an issue. Therefore, this study aimed to discriminate external defects in soybeans using fluorescent images in addition to color images, using a model that is less expensive than deep learning. Color images and fluorescence images of soybeans at an excitation wavelength of 365 nm were taken, and visually labeled into six categories: normal, wrinkled, peeled, pests, denatured, and insect-damaged grains. Image features were extracted for both color and fluorescence images, and classification models were constructed using Support Vector Machine (SVM) with three input patterns: (a) color image features alone, (b) fluorescence image features alone, and (c) color and fluorescence image features input. The test accuracy was 75.06%, 58.28%, and 76.91%, respectively. In addition, two and four classifications were devised in order to anticipate the needs of the field. The accuracy reached 82.06% and 94.99% for four category and two category simultaneous input of color and fluorescent images, respectively. Especially in the two categories, the discrimination accuracy exceeded 90%, indicating that a highly accurate discrimination model could be created. Furthermore, as a result of visualizing the features important for discrimination using the ablation study, it was found that fluorescent images were effective in addition to color images for discrimination in the six categories. The importance of shape information, namely, perimeter and circularity, was high in the discrimination model for all categories, indicating that shape information is the most important information for discriminating external defects in soybeans. From those results, it can be concluded that the combination of conventional color images and fluorescence images is effective for classifying soybean external defects.

  • Kenichi Kusunoki, Naoki Ishitsu, Toru Adachi, Osamu Suzuki, Ken-Ichiro ...
    2024 Volume 5 Issue 2 Pages 22-39
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    This paper describes the development of the “Collaboration with Startups for Localized Severe Weather Countermeasures: Building a Real-time Disaster Prevention Field using AI” project, which is positioned as part of the BRIDGE program. The core technology used in this BRIDGE project is based on collaborative work between the Meteorological Research Institute and East Japan Railway Company, which involves using deep learning to automatically detect low-level rotational airflows associated with wind gusts in high-resolution radar data. The key objectives of the BRIDGE project are to enhance the existing deep learning models for accurate tornado vortex detection, expand the application scope beyond railway operations to broader sectors by integrating GPS location data, and foster industry- academia-government collaboration, including partnerships with startups, for efficient technology development and practical implementation. The paper outlines the principles of observing tornado vortices using Doppler radar, the construction of deep learning models for detecting tornado vortex patterns, and the processing flow and application examples in train operation control, building upon our previous work presented in Kusunoki et al. (2022). It also provides an overview of the BRIDGE program and the positioning of this BRIDGE project within it, highlighting the industry- academia-government collaboration system and the involvement of startup companies. The initial results of the project are presented, including the development of advanced deep learning models for tornado vortex detection, comparing their performance against the VGG model which we previously developed, and the efforts towards building a real-time disaster prevention information dissemination system integrated with GPS data. The paper concludes by discussing the expected future developments, academic insights, and societal impacts of this research, which aims to strengthen resilience against localized meteorological disasters while contributing to the advancement of tornado research.

  • Young Kwang Hwang
    2024 Volume 5 Issue 2 Pages 40-44
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    In this study, the smoothed particle hydrodynamics (SPH) code for solid mechanics is extended to simulate the three-dimensional (3-D) behaviors of reinforced concrete (RC) structures. The reinforcing bar is discretized and modeled as a 3-D truss element. To save the total degrees of freedom in the system domain, an implicit representation of the reinforcement is introduced, where the displacements in the reinforcement are interpolated using the displacements of adjacent SPH particles representing the matrix material. An isotropic damage model is introduced to reproduce the concrete damage behaviors. Thereafter, these numerical methods of the SPH, reinforcement model, and the damage model are integrated and validated through simulations of tension stiffening behaviors of a RC member. It is demonstrated that the proposed approach effectively predicts the longitudinal force and strain curve, as well as the multiple cracking behaviors of the RC specimen.

  • Yoshiyuki Yamamoto
    2024 Volume 5 Issue 2 Pages 45-56
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    This study investigates the enhancement of deep learning-based roadside vegetation detection using vegetation indices. Vegetation detection is crucial for maintaining clear sightlines for autonomous driving and as an indicator of potential drainage issues and road deterioration. A Faster R-CNN architecture was employed to analyze images from vehicle-mounted cameras, with three separate input types evaluated: standard RGB images, Excess Green Index (ExG) images, and Color Index of Vegetation Extraction (CIVE) images. In addition, an integrated approach was developed that combined these input types. The results demonstrate that the integrated approach consistently outperformed individual input-based detections, achieving the highest Average Precision (AP) in both validation and test datasets. CIVE-based detection showed the highest overall performance among single input types, particularly in the test dataset. Vegetation indices generally improved detection accuracy compared to the standard RGB input, especially for challenging scenarios. However, all input types struggled with small object detection, indicating an area for future improvement. The study also revealed varying levels of detection consistency, with RGB-based detection showing the highest consistency across data sets. These findings contribute to the advancement of roadside vegetation detection techniques and suggest potential applications in comprehensive road condition assessment, automated maintenance planning, and early detection of drainage problems, complementing existing crack and pothole detection methods.

  • Zhenyu Yang, Hideomi Gokon
    2024 Volume 5 Issue 2 Pages 57-65
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    Understanding intercity transportation demand and trends during snowstorms is crucial for mitigating traffic accidents and congestion. This study analyzed the characteristics of intercity transportation demand using Agoop mobile GPS data, focusing on the 2018 Fukui Heavy Snow Event. The K-dimensional Tree (KDTree) algorithm for neighbor matching was employed to examine the spatiotemporal characteristics of intercity transportation demand, providing insights into the overall trends under the influence of the snow disaster. Additionally, Singular Value Decomposition (SVD) was utilized to decompose and reduce the dimensions of the spatiotemporal OD matrix, facilitating the identification of the composition of intercity transportation flow. The study identified three phases of the event: the stable phase (January 27 to February 2), the snow disaster phase (February 3 to February 11), and the recovery phase (February 12 to February 16). During the snow disaster, intercity transportation demand dropped by 67.86% compared to the stable phase. Intercity transportation demand during the snow disaster included daily demand (M1) and special demand (M2). M1 traffic primarily originated from Fukui City, Sabae City, Awara City, and Sakai City, with Fukui City and Sakai City being key points of departure and arrival. In the M2 model, traffic from southern to northern Fukui Prefecture nearly ceased, shifting from a longitudinal pattern along National Route 8 and the Hokuriku Shinkansen to a horizontal distribution towards Ono and Katsuyama. Departure flows from Sakai City decreased significantly, while Tsuruga City saw a significant increase in traffic. This indicates that the snow disaster mainly affected northern Fukui Prefecture, especially Fukui City and Sakai City.

  • Zicheng Han, Suguru Kodaka, Kazutoshi Nagata, Yuina Ota, Kunitomo Sugi ...
    2024 Volume 5 Issue 2 Pages 66-73
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    Since the 1995 Hyogo-ken Nanbu Earthquake, rubber dampers have been increasingly adopted to improve earthquake resistance. However, in recent years, a deterioration phenomenon has been reported in which the vulcanization bonding area between the bottom of the rubber damper laminate and the bottom steel plate peels off. Although this delamination causes a significant reduction in seismic performance, a method for accurately measuring the degree of delamination has not yet been established. Previous studies have shown that there is a relationship between the degree of rubber damper delamination and the shape of warpage. Therefore, in this study, we have learned this relationship by deep learning using a neural network and developed a system to estimate the degree of delamination from the shape of the rubber damper’s warpage. As a result, it was found that deep learning can be successfully used to estimate the degree of delamination from the shape of the warpage of a rubber damper.

  • Hailong Wang
    2024 Volume 5 Issue 2 Pages 74-86
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    This paper presents an isotropic-confining triaxial system particularly designed for three-dimensional (3-D) volume change measurement of unsaturated compacted bentonite when being wetted. The specimen size was set to be small (diameter less than 30 mm and height of 10 mm) to reduce testing duration, which, as a result, introduces many difficulties to design the apparatus and to measure volume change with sufficient accuracy. In the triaxial system, two volume change measurement methods were applied, local deformation measurement by using laser displacement transducers (LDTs) and global measurement method with a differential pressure transducer (DPT) to monitor volume change of cell water. Another two methods to measure final volume change, the final water content method and the pycnometer method, were also newly proposed. Technical details to improve measurement accuracy were described. Testing results suggest that the measurement accuracy in terms of absolute volume change variation was of 0.09-0.12 ml, which is better than many currently available techniques.

  • Shijun PAN, Keisuke YOSHIDA, Yuki YAMADA, Takashi KOJIMA
    2024 Volume 5 Issue 2 Pages 87-94
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    At present, the Ministry of Land, Infrastructure, Transport, and Tourism (i.e., MLIT) has installed about 3,000 CCTV cameras in rivers throughout Japan. Initially, these devices were installed to monitor water levels during floods and visually inspect damage after earthquakes. Moreover, the riverine environment management researchers (i.e., “the researchers” in the following content) have been trying to survey the situation of rivers and embankments through patrolling. In addition, night-time patrols are carried out only in certain areas at certain times of the year for certain purpose. It is difficult for the researchers to monitor the status of riverine areas 24 hours a day. In recent years, researchers have conducted river space utilization surveys (i.e., RSUS) using camera images and machine learning. However, in the case of nocturnal RSUS, there is no detailed research on this topic. Therefore, the authors of this study (i.e., “the authors” in the following content) aim to understand the number and behavior of humans using river space at night, and confirm the performance of the models depending on the brightness difference in the images.

  • Masato Abe, Koichi Sugisaki, Pang-jo Chun
    2024 Volume 5 Issue 2 Pages 95-105
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    Application of AI(Artificial Intelligence) technology in infrastructure management is advancing. This article reviews recent advances and state of the art of the field from articles published in Japanese journal of “Artificial Intelligence and Data Science” issued by JSCE(Japan Society of Civil Engineers), with emphasis on use cases in image recognition, language models, prediction, acoustic data, monitoring, nondestructive evaluation, and combination with physics. Open data related to inspection record are pushing the progress of data driven approach in Japan. Inspection support by image data and interpretation of natural language are in progress by development LLM(Large Language Model). Structural monitoring is advancing by introduction of machine learning and CNN(Convolutional Neural Networks) technologies. Combination with physics is expected to improve clarity and quantification of AI.

  • Jiaming LIU, Kai XUE, Boyu ZHAO, Mayu HAZAMA
    2024 Volume 5 Issue 2 Pages 106-115
    Published: 2024
    Released on J-STAGE: November 20, 2024
    JOURNAL FREE ACCESS FULL-TEXT HTML

    This study presents a novel method for evaluating the stripping ratio of road lane markings using a two-step deep learning-based semantic segmentation approach. In the first step, a segmentation model identifies the lane area on the road under various conditions. The second step involves using a deep learning-based segmentation model, which processes the collected training data and annotates it using a multi-section binarization method with various data augmentation strategies to distinguish between stripping and non-stripping areas within the lane markings. Unlike traditional methods that rely heavily on manual inspection or low-robustness image processing techniques, this approach leverages smartphone cameras mounted on moving vehicles to capture and automatically analyze the stripping ratio of lane markings with high accuracy across entire road segments. The results demonstrate a high correlation (R² = 0.9827) with manual evaluations, highlighting the potential of this technique to significantly reduce labor-intensive assessments. The efficiency and effectiveness of this method could revolutionize road maintenance by providing reliable, rapid, and cost-effective assessments of road markings.

feedback
Top