Gi-Jun Lee

Reliable in-situ strength information is critical for construction-stage decisions, yet conventional UPV–UCS correlations are typically established on companion cylinders, which may overestimate the state of cast reinforced members. This study addresses this gap by developing an IoT-enabled embedded ultrasonic sensing system that tracks age-dependent UPV along a fixed internal path (200 mm) in cast reinforced concrete from casting to 28 days, and by quantifying the persistent discrepancy between embedded-path and cylinder-based surface UPV under identical mix designs. Across all mixes and ages, embedded-path UPV remained lower than surface UPV (ΔUPV = 118–415 m/s), suggesting that cylinder-based UPV may provide unconservative strength inference for cast members under the investigated configuration. By pooling embedded UPV with companion-cylinder UCS results, we establish a conservative empirical UCS–UPV model (UCS = 0.3846e0.0011V_p, R² = 0.82) with uncertainty metrics (RMSE and MAE). A cable-free LoRa–Wi-Fi architecture with gateway integrity checks and REST-based cloud upload demonstrates the feasibility of periodic near-real-time visualization for curing-stage monitoring, although full communication-performance benchmarking was beyond the scope of this study. Although direct member-strength verification of the cast specimens was not performed, the results support embedded-path UPV as a practical complementary indicator for construction-stage decision making.

Ground settlement induced by shield tunnel boring machine (TBM) excavation is a major geotechnical concern in urban tunneling because it may affect the safety of adjacent structures and underground infrastructure. In this study, machine learning models were developed to predict the maximum settlement induced by shield TBM excavation using a three-dimensional numerical analysis database comprising 320 simulation cases generated from combinations of tunnel diameter (D), ground elastic modulus (E), face pressure (FP), and backfill pressure (BP). Random forest (RF) and extreme gradient boosting (XGBoost) models were developed and compared with an existing regression-based settlement prediction equation. Predictive performance and generalization capability were evaluated using random split and GroupKFold validation techniques. Under random split validation, RF achieved the highest predictive performance, with a coefficient of determination of 0.997 and a root mean square error of 0.438 mm, followed by XGBoost. Both machine learning models outperformed the existing settlement prediction equation. However, model performance decreased substantially under GroupKFold validation, indicating limited generalization capability under unseen D–E grouped conditions. The results demonstrate that the developed machine learning models provide accurate predictions within the range of tunnel–ground conditions represented by the adopted numerical analysis database. The findings highlight the importance of evaluating both predictive performance and generalization capability, particularly when machine learning models developed from numerical analysis databases are applied beyond the conditions represented in the training database.

A utility tunnel is a complex underground facility that serves as a critical infrastructure integrating and operating systems such as electricity, telecommunications, and drainage within a city. However, various hazard events such as fire, flooding, condensation, damage, and intrusion can occur within utility tunnels, posing risks not only to the degradation of facility functions but also to potential human casualties and economic losses. Therefore, it is crucial to establish prompt and effective response measures for these hazard events. Unlike previous studies that focused on individual hazard types, this study proposes an integrated and standardized multihazard response framework for utility tunnels. Through case analysis, fire, flooding, condensation, damage, and intrusion were defined as representative hazard events, and the response stages for each were classified into attention, caution, warning, and critical levels. Appropriate response scenarios were developed for each stage, providing prompt and efficient response measures tailored to different situations. The proposed framework offers a unique contribution by presenting a unified structure that supports stage-based management and can be directly applied to smart monitoring and control systems in underground infrastructure. This study is expected to contribute to improving the disaster prevention and response systems of utility tunnels and enhancing overall facility safety.