In recent years, many reports of damage caused by debris flows have been published. However, sediment disaster hazard areas are designated only in areas where there is a risk of human damage, and the risk of damage throughout the country cannot be exhaustively evaluated. In this study, we developed a method to estimate the location of slope failure and distance of debris flow occurrence along a mountain stream using deep learning, which is computationally inexpensive, and to assess the risk of damage from debris flow. As a result, a hierarchical neural network model was constructed by learning the topographical information at the collapse site, and it enabled classification of the collapse with an F value of over 0.7. The LSTM model, which learned the traces of debris flow and topographic information along the stream, enabled easy estimation of the probability of debris flow arrival along the mountain streams, including the uncertainty of the streambed condition and the physical state of the generated debris flow. Furthermore, our results indicate that the combination of these methods can be used to exhaustively assess the risk of debris flow damage over a wide area.

 It is required that the stakeholders work together to improve the balance between flow capacity and basin storage so that the flood control of the entire basin is increased. Therefore, it is necessary to analyze and visualize the relationship between water storage, infiltration, and runoff in a basin during heavy rainfall events. In this study, a coupling method of rainfall-runoff and flood flow analysis was developed to analyze flooding phenomena with rainfall as a given condition. Basin Water Balance Map showing the relationship between water storage, infiltration, and runoff in a basin is developed, and its application to flood control in a basin is examined. Specifically, we developed an analysis method that combines a tank model, which is a rainfall-runoff analysis method, and flood flow analysis, and applied it to the Toyohira River to demonstrate the usefulness of this analysis method and to show how the Basin Water Balance Map can be used.

 Flood risk is projected to escalate due to the intensification of heavy rainfall associated with climate change, underscoring the growing importance of revisiting settlement guidance and urban planning based on potential flood inundation in addition to structural countermeasures. However, in Japan, the strong land ownership system complicates the implementation of land-use regulations, thereby increasing the need to promote settlement in low-risk areas. Previous studies developed an agent-based intra-regional relocation model focusing on the Kyoto Basin to support such planning efforts. Building upon this, the present study introduces a method to incorporate non-stationary demographic and urban changes including total population growth, the introduction of new railway stations, and new town developments from 2000 to 2015— as boundary conditions within the model. As a result, by explicitly accounting for these temporal dynamics, the model successfully reproduced the past trends in flood-exposed households and the temporal changes in flood-induced economic damage. While the overall increase in flood-exposed households across the Kyoto Basin was limited, areas with new railway stations and new town developments showed a relatively higher increase in flood-exposed households compared to the entire basin.

 The shortage of truck drivers in the logistics industry has become increasingly serious due to avoiding overly long working hours, and there is an urgent need to improve the working environment. The introduction of towing cargo vehicles using electronic-connected technology is expected to be a solution to these problems, but various issues remain to be solved, such as the development of facilities such as formation bases and the establishment of driving operation methods. In order to establish a safe and smooth operation method for long strings of electronically-connected vehicles, this study focuses on situations where such vehicles merge with other vehicles near expressway on-ramps, which is considered to have a particularly large impact on the safety and smoothness of traffic flow. This study analyzed the impacts of electronically-connected vehicles on surrounding traffic near on-ramps when they merged, and the effects of measures to improve the safety and smoothness of traffic by using an analysis system developed by combining existing traffic simulation models.

 To simplify complicated reinforcement arrangement at knee joints in RC box structures, we have developed the corner reinforced unit in which the main bars in walls are connected at the unit. A calculation method for the ultimate capacity at the knee joints was developed based on failure mechanisms of the conner reinforced unit. Comparison of the calculated values with previous experimental results shows that the method proposed can estimate the ultimate capacity. Besides a design flow for the corner reinforced unit was proposed. Finally, a design examination for an actual road box culvert based on the design flow was carried out to confirm the validity of the design flow proposed.

 A method to derive a bond stress-slip relationship between FRP sheet and concrete based on a debonding mechanism, instead of using the experimental bond stress-slip relationship obtained from bond tests of FRP sheets bonded to concrete, was developed. This bond model expresses the debonding behavior of the FRP sheet as a saw blade failure of the concrete surface layer and is derived from the equilibrium of forces, deformation conditions, energy equivalence, and constitutive laws of concrete and resin. Numerical analyses of bond tests with the developed bond model show good agreement with experimental results, irrespective of sheet stiffness, bond length, and resin type. The proposed model was also applied to existing bond tests where the debonding depth was unknown, and was as accurate as the existing bond models. Furthermore, as an example of using the proposed model, the influence of material properties was evaluated, and the proposed model was applied to pulling from both ends with different boundary conditions to demonstrate the model’s validity.

 Efforts are being made to use cross-laminated timber (CLT) in the civil engineering field, where there have been almost no previous examples. Against this backdrop, we considered the use of CLT as a countermeasure against soft ground. To investigate its feasibility, a demonstration experiment was conducted in which CLT was laid horizontally on actual soft ground to reinforce the soft ground and a full-scale embankment was constructed on top of the CLT. In this experiment, we investigated the workability and economic efficiency of the construction, differential settlement and stability, and soundness of the CLT after approximately 650 days. The effect of this countermeasure on differential settlement and stability was not clear, but the results showed that the CLT at least did not have a negative mechanical effect, and it had good trafficability, with no biodeterioration or mechanical deterioration. Therefore, this application of CLT was feasible, although verification of its cost and mechanics is an issue to be addressed in the next phase of this research.

 This study focuses on the remote operation of construction machinery using close direct visual information, which is utilized during emergency response just after a disaster. An experiment was conducted to compare the task completion times between onboard and remote operations. In the experiment, operators were tasked with using a hydraulic excavator to remove randomly placed solid objects from a designated area, and their task completion times were measured, while video footage of the operations was recorded. Quantitative analysis of the obtained time duration data confirmed that remote operation takes significantly longer time than onboard operation. The primary reason for this difference was found to be the positioning operation conducted prior to the working operation, particularly when moving individual solid objects. This finding, which has not been previously highlighted, suggests that future improvements in remote operation efficiency should focus on enhancing the operational environment and the user interface, with special attention to the positioning operation.