In a large-scale disaster, supporting systems for vulnerable people is important for the minimization of personal suffering. Everyday relationship leads to the refuge support at the time of disaster. On the other hand, supporting systems for vulnerable people has problem. For example, it is not consider many supporters are not in weekday daytime.
 This paper diagnose about supporting systems for vulnerable people in Tonaki-village, Okinawa which is a depopulated aging island region in consideration of them. In addition, it considers what kind of supporting systems is required, while maintaining a landscape and not considers the measure by an infrastructure.
 This study estimate cognitive relationships between inhabitants and road blockade as local diagnosis. As a result, although the depopulated region generally has strong connection between the neighborhoods, it was shown that is not necessarily right. It proposes six methodology of matching of supporters and vulnerable people considering cognitive relationship between neighborhoods and unsupportable probability in weekday daytime and night-holiday. This is applied to Tonaki-village. And the difference in matching result of supporters and vulnerable people in two situations from the time index etc. was evaluated quantitatively. As a result, in Tonaki-village, all vulnerable people can evacuate if there was a cognitive relation to public health nurse and supporters, and it matched in the optimal combination such as MinSum and MinMax.

View full abstract

 Deaths resulting from falling accidents account for 40% of all labor-related deaths in the construction industry and outnumber deaths from any other cause at construction work sites. This paper presents the current data on accidental falls from “cliffs and slopes, ” the most common places from which falls occur on construction work sites in the civil engineering industry, and a survey and analysis of labor accidents (fatal accidents) performed to obtain basic data useful for establishing measures to prevent falling accidents. The results of the analysis were used to conduct a factor analysis with the m-SHEL model, an analytical approach in human factor engineering, in order to study necessary items and technology for preventing falling accidents during construction works on slopes.

View full abstract

 Inundation has been occurred in Japanese urban areas. Simulations are often used to assess pluvial flood risk, but such approaches tend to be highly complicated and cannot evaluate detailed topography. Therefore, this study aims to analyze the characteristics of flooded areas and make a potential map for pluvial flooding using high resolution data and machine learning. We extracted some indices related to roads and topography, and they were analyzed using Random Forest. The study areas in Tokyo were classified into flooded and non-flooded area with accuracies about 80-90%.

View full abstract

 In recent years, enormous damage has been caused by natural disasters. Disaster Medical Assistance Team (DMAT) play an important role in minimizing human suffering. However, since factual information is intermingled with disinformation in the event of a disaster, it is difficult to obtain correct information, such as where and to what extent medical care is required, in order to dispatch DMAT appropriately. In this study, using a victims estimation formula empirically derived from the cumulative death toll data with the passage of time in past seismic hazard, an optimization problem was addressed to minimize the number of deaths.
 Consequently, it was found that early dispatch of DMAT would be effective, and DMAT should be dispatched according to the circumstances (number of DMAT, conditions of affected areas) in order to minimize the cumulative death toll. And, by setting the conditions of the affected area, it was possible to consider the way of reasonable DMAT dispatch. Thus, it has become possible to suggest the way of effective dispatch of finite DMAT, to use limited DMAT resources efficiently, and to help make decisions.

View full abstract

 Decision makers must decide the disaster correspondence by using uncertain information. When the decision maker makes this decision, they have the dilemma to prioritize which of “the rapidity of correspondence” and “the accuracy of information”. The result for disaster varies according to the difference in priority.
 This study approximate the change of cumulative number of death with time lapse by using the Weibull distribution and proposed the Bayesian estimation model to predict the total number of deaths by utilizing the change of cumulative number of deaths. By considering the estimated total number of deaths with past earthquakes cases, it was possible to estimate the time period to make the disaster response decision. Therefore, this would be a judgment material to reduce the dilemma of decision makers.

View full abstract

 In this paper, we consider the transition process of a disaster prevention consciousness and action as a disaster subculture, and present an analytical framework using a mathematical simulation in order to enable quantitative consideration. After an unknown parameters estimation by using a questionnaire survey data, we describe the actual situation in this simulation model, and evaluate an influences imparted to the population ratio of four states related with the disaster prevention consciousness and action by some explanatory variables such as a force of contagion, a move-out ratio, and so on. As a result, we receive valuable suggestions that it is impossible to avoid a fading of the disaster prevention consciousness, and that it is possible to vitalize the disaster prevention action continuously if some social conditions are satisfied.

View full abstract