Withdrawal of public transport caused by population decline and motorization, mainly in rural areas, presents many difficulties. As a countermeasure, a “mobility policy” to strengthen the network centering on public transport is regarded as important. Among mobility policies, a review of the plan must be conducted for both a location normalization plan and a regional public transport network formation plan. However, the relation between the formulation status and various attributes of the municipality is insufficiently described in the related literature. Therefore, this study was conducted for quantitative analyzes of the relation between the state of development of mobility policy and local governments attributes. Results clarify that local government staff awareness such as existence of a Department of City Planning system and effort status of various mobility policies for local government staff recognition influence the plan formulation.

This study develops a deep learning model for forecasting pavement deterioration in an expressway. Road pavement plays an important role for road traffic by providing comfortable and safe driving environment. It is therefore crucial to forecast its deterioration, based on which timely and strategical maintenance can be implemented. Much literature has been reported on predicting pavement deterioration based on statistical modeling approached by utilizing pavement surface condition data. However, few studies have been conducted based on the deep learning approaches, which could be suitable in modeling the nonlinear process of the pavement deterioration. This study employs the multilayer perception network for developing the pavement deterioration forecasting model in an expressway network of Shikoku, Japan. The proposed model is then compared with a benchmark model which is based on the multivariable regression model. The comparison by the RMSE has identified that the proposed model is superior in forecasting the pavement deterioration. It has also been revealed that the model has limited capability in migrating to different road networks due to the overfitting.

In order to maintain local public transport whose users are decreasing, it is important to gain not only cooperation of local government and transport operators, but also engagement of local stakeholders. For the promotion of a wide spectrum of engagement, public transport should be accountable to the local stakeholders by visualizing and verifying its social impact. In this paper, we tried to apply a social impact assessment using the PDARU cycle presented by the Japanese Cabinet Office for public transport. By conducting the social impact assessment for public transport in Shodoshima, we assessed the social impact that public transport brings, and verified its structure.

In this study, it focused on getting on and off autonomous vehicles in urban areas and quantitatively grasped the influence on traffic. In this analysis, it simulated parking on a single road by changing the parking frequency and the shape of the curve side. From these results, it was confirmed that in the case of the on-street parking type without a stop zone, the travel speed is significantly reduced when the parking frequency is high, and the delay time is increased. In addition, it was found that even if the stop zone is provided, the longer the getting on and off section, the more likely the opportunity for deceleration occurs. From the above results, it was found that it is important to control the parking on the road by limiting the getting on and off sections when introducing an autonomous vehicle in an urban area.

This study is intended to find out the high-density area where the freight vehicles come for loading and unloading. Two different types of data, the digital tachograph data of freight vehicles which were collected from Urban Freight Survey in Tokyo Metropolitan area and the on-street parking data which was collected by Tokyo Metropolitan Police Department, were combined on the GIS software and analyzed the relation of these two data on each block. And then, the selected area can be classified into eight patterns and additional analysis was implemented in the districts with classification I, II, and III. As the result, the characteristics of freight concentrated districts were different in each classification. Hence, it was found that the countermeasure should be thought based on each classification characteristics.

In regional economic activities, the stock effect brought by road infrastructure may not always follow the fundamental economic theory since the technological innovation would surely occur after the investment. Therefore, finding on causal relationship between the actual regional characteristics is the big issue to understand how infrastructure stock effects appear. In the past studies, the causal structure of the variables was assumed on the basis of economic insights and validity of economic theory. In this study, the causal relationship between improvement of accessibility by road improvement and economic growth is found by the statistical characteristics of the data, called causal discovery. Causal discovery requires no assumption on causal structure, but purely investigate the independence among the variables. The proposed procedure in causal discovery is applied to confirm the validity.

The Ministry of Land, Infrastructure and Transport says it is important to prevent and preserve maintenance for bridges that are rapidly aging. Periodic inspection is required for all bridges, but Proximity visual inspection currently being done is not efficient because money and people do not have enough. Many alternative systems are being developed. In this research, we propose “Image Visual Diagnostic System” which diagnoses damage by looking at images taken by a super high resolution camera. In this issue, it is clear about “the influence that inspection skills have on the diagnosis result” which is a problem that occurred when proposing the system and “the correlation between inspection skills and time required for diagnosis” necessary for more efficient inspection I made it.

In winter Hokkaido, road visibility information is necessary to prevent risk of low visibility on the road due to snowstorm caused by low pressure. Therefore, using fixed camera on the national road which is analyzed by WIPS can detect low visibility from images. However, using fixed camera to detect low visibility are different from driver's gaze and hard to know visibility information where there is no fixed camera. Thus, we analyzed the images taken by on-board video camera during a snowstorm and verified whether WIPS can evaluate visibility information for road management. At first, we conducted an experiment which evaluate driver's visibility and ease of driving to know relationship between WIPS and driver's evaluation using on-board video camera. As a result of experiment, WIPS of on-board video camera can be similar to evaluation of driver's visibility. Next, we analyzed visibility information of fixed camera and the on-board video camera in the national route 40 from Toyotomi to Horonobe. As a result of analysis, we can find section where tend to occur low visibility and is valid in generating temporal and spatially dense visibility information that is difficult with fixed camera.

In this study, in order to gain insights into changes in the observance of traffic directions, we compared the traffic directions of the bicycle participants in bicycle accidents in Chiba Prefecture between 2007 and 2017. In a bicycle accident that the car went straight, the percentage of bicycles traveling on the left side of the road increased from 74.2% to 86.6%. In addition, there was a 38.7% reduction in the number of accidents that occurred when a car entered a road from a parking lot, which is an accident in which bicycles passing on the right are likely to be shared.

Road Traffic Act was amended on June in 2008, four-wheeled vehicle driver must drive to make rear seat occupant fasten seatbelt in Japan. On October in 2008, enforcement not to fasten rear seatbelt had started only on expressway. So, according to seatbelt usage social survey on road side carried out by Japan Automobile Federation (JAF) and National Police Agency (NPA), rear seatbelt usage rate had increased more on expressway and surface streets just after Road Traffic Act was amended. It was estimated to reduce fatality and severely injury rate on rear seat with rising rear seatbelt use rate. And it was calculated that about 500 fatalities and severely injuries on rear seat had decreased. To calculate attribute of rear seatbelt use lower rate, the rates on accidents of young age and male are low a little. Rear seatbelt use rates on accidents in taxi, hired car, heavy vehicle and the rates of which driver does not fasten seatbelt is very low.

In order to realize effective countermeasures for congestion, it is required to detect signs leading to congestion and prevent them. However, although there are many research cases that tried to detect a sign of congestion, its accuracy is not high. This paper examines whether traffic flow analysis by sonification is a new method leading to high congestion prediction. The speed of the two points of the bottleneck part and its upstream part was converted into frequency change and the traffic volume was converted to volume change. In evaluation experiments, it was evaluated whether subjects were able to distinguish audible data of free stream, critical flow, critical flow before congestion, congestion flow. As a result, it turned out that the subject distinguished and impressed the sounds of free flow, critical flow, and congested flow. In addition, it was found that audible data is easier to distinguish traffic flow data immediately before congestion than visualized data.

There has been an increase in the cases of installation of the Pace-maker light (hereafter, PML) as a countermeasure against traffic congestion at sag bottlenecks of freeways in Japan. PML creates a flow of LED light traveling with constant speed alongside the car. It intends drivers to perceive the speed gap between PML and their own vehicles and encourage the spontaneous speed recovery on uphill section at sag. It has been reported that traffic congestion at some bottlenecks were mitigated due to the installation of PML, though its influence on the driving behavior and the mechanism of congestion-mitigation were still unrevealed. In this study, based on a car-following experiment conducted on the Hanshin Expressway Route 3. From the obtained data, in the first, we analyzed quantitatively the relative speed by PML operation conditions. Next, the car following-behavior model was constructed in consideration of the deviation of the relative speed due to the driver's cognition and then the parameters were estimated. As a result, by operating PML, 1) from the quantitative analysis, PML contributes to the stabilization of following behavior in the whole driver. 2) From the analysis by the car following-behavior model, it is confirmed that the parameters related to the cognition of the relative speed change are improved and the passing time of the target section of the car-following group is shortened by running simulation.

Since lane changing behavior is interdependent behavior among drivers, modeling based on game theory is useful. However, the historical model covers only the gaps that face, there is no model yet to consider the selection behavior of multiple gaps. Therefore, in this study, we modeled by choosing multiple gaps as a two - step game and confirmed that the proposed model has a higher explanatory power than the past model by analysis using simulation data.

The demand of aviation in the world increases year by year, and it is an important project that Haneda airport wants to strengthen the function of it. Moreover, there are many foreign countries introduced a new operation system which can improve the air traffic management. Because of this system, those countries get many benefits, such as reducing the workload of controller, reducing the consumption of fuel and the environmental load, and so forth.

In this paper, the goal is to created a new operation system that is suitable for Haneda Airport, and to analyze the imported effect of this system in the sky of Tokyo. So, I developed an simulator in airspace that can quantitatively evaluate the imported effects of a new operation system. In addition, I created a new operation system which is used in the sky of Tokyo, and analyzed imported effect of it. And finally, the imported effect of it from the analysis results was examined.

This research analyzes the tendency of cognitive bias perceptions for traffic accident risk on the Matsuyama urban area considering residential street. This research conducted a questionnaire survey for the drivers living in Matsuyama city area to understand not only situation of residential street utilization, but also intention of main road and perception to traffic accident risk. It is to analyze how drivers recognize the traffic accident risk for both main road and residential street in Matsuyama city. The analysis results from questionnaire survey data describes that 40 – 50% of drivers currently recognized traffic accident risk which are risk to cause, risk of getting caught and fatal accident rate. In addition to that drivers who is thinking that they can avoid the risk to cause traffic accident on residential street and the risk of getting caught tends to overly evaluate sever accident rate on main road by examining estimated parameters from ordered response probit model.

In this research, when setting up a two-stage crossing facility with no signal in a basic road section, in order to clarify the traffic volume of the road where the smoothness is not lost, various basic road conditions 1 cases set in advance are used. By carrying out traffic flow simulation with traffic volume patterns, it was calculated how much delay time would occur by installing a two-stage crossing facility. In addition, we calculated the delay time under various traffic conditions, and summarized the traffic conditions that affect the delay time and their degree. As a result, the general delay time that may occur on the road where the no-signal two-stage crossing facility is revealed. Additionally, the influence and extent of the traffic conditions such as vehicle speed, large-sized vehicles and vehicle group formation on delay time are clarified.

This study analyzed the relationship between vehicle movements and intersection geometries at signalized intersections by field observations. For left-turn traffic, it is found that the approaching speeds become higher with the longer setback distance of stop-line and the larger curb radius induces the higher speed inside of intersections. In addition, it is revealed that the channelization for left-turn vehicles by zebra marks along curb corner become a cause of the wider vehicles' trajectories and is ineffective in reducing the turning speed at intersections. On the other hand, for through traffic, it is also clarified that the longer distance between stop lines at intersection is, the more risky movements such as over speeding and rushing into intersection during inter-green periods occurs. It is summarized as follows. Shortening of corner radius has an effect on reducing the speeds of left-turn vehicles at intersections and the shortening of opposing stop lines are expected to reduce the risky behaviors of through traffic during inter green period.

In order to promote crossing pedestrian accident measures in Japan, the purpose of this study is to clarify the effective and safe cross-sectional configuration of a simple two-stage crossing facility that can be installed even on roads with no signal and a sufficient width. Traffic experiments of two-stage crossing facilities with different patterns of cross-sectional configuration were conducted, and their effectiveness and safety as traffic safety measures were investigated. As a result, on the roads surveyed in this study, the widths of the roadway section are 7m to 9m, if the way of using the road is limited as necessary, the possibility of installing two-stage crossing facilities that could be used effectively and safely was confirmed.

When taking measures for road traffic accidents, it’s desirable to focus on each measure point for example road or intersection. However; that measures is inefficient and difficult to consider of structure of road network as macroscopic factor. The objective of this study is to quantitative evaluation of structure of road network and estimated traffic accidents based on that. The study method is calculating of the ratio of traffic capacity on each direction and plotting them on polar chart, and then we quantitative evaluate that by using standard deviation. After that, we categorize sample by them using cluster analysis and, analyze accidents factors using Poisson distribution model. As a result, we can categorize by share ratio of arterial road capacity and, for decreasing accidents we needed not only improvement arterial road extension but also them direction.

Recently, the system that improves the convenience of users has been spreading. Even in the field of public transportation, attention is paid to utilization of traffic big data. In this research, we have grasped actual traveling situation and usage situation from location data and loading survey data. As a result, it clarified from both use and operation such as delay section and connecting. Next, timetables were created using the average value from the distribution of the time required between bus stops, and timetables were created using expressions of timetable optimization using operation data. We compare these timetables and considered optimum timetable of community bus. As a result, it became clear that the timetable using the average value is appropriate.