Archives Search

Abstract

On toll roads where tolls vary based on the existence of towed vehicles and trailers, the ability to determine the number of axles become a necessity in automating the vehicle classification process. Traditional systems have utilized pressure sensitive sensors buried unde...Lead moreOn toll roads where tolls vary based on the existence of towed vehicles and trailers, the ability to determine the number of axles become a necessity in automating the vehicle classification process. Traditional systems have utilized pressure sensitive sensors buried underground to determine the number of axles but this requires large costs of installation and maintenance which became a disadvantage. Attempts to utilize contactless methods to determine the number of axles have had issues with accuracy. Recent advances in the area of deep learning technologies have made improvements in accuracy possible. Here is a presentation on a pilot to determine the number of vehicle axles utilizing deep learning where vehicle side view data obtained from laser scanner based vehicle detection system is entered as data.

Abstract

The days towards the realisation of fully driverless Autonomous Vehicles (AVs) on the roads is no longer a distant future and cities around the world are racing to be the first to realise this dream. With artificial intelligence (AI) gaining traction in recent years and t...Lead moreThe days towards the realisation of fully driverless Autonomous Vehicles (AVs) on the roads is no longer a distant future and cities around the world are racing to be the first to realise this dream. With artificial intelligence (AI) gaining traction in recent years and the increasing interest to roll out AVs on the roads along with existing human-driven vehicles, questions remain on how safe are AVs and how well they can co-exist and interact with existing road infrastructure and human-driven vehicles. To answer these questions, an AV Monitoring and Evaluation System has been developed for the purpose of monitoring and evaluation of the AV trials conducted in Singapore. This paper covers the details on how the Land Transport Authority (LTA) of Singapore, in realising the Singapore Autonomous Vehicle Initiative, leverages on advanced ITS technologies in monitoring and evaluating the readiness of AV technologies in Singapore.

Abstract

In Japan, trucks are used for most of the logistics operation, which accounts for 90% of Japan’s transportation in terms of tonnage. At present, logistics companies are facing an increase in fuel and labor costs as well as competition for transportation fees, and efficien...Lead moreIn Japan, trucks are used for most of the logistics operation, which accounts for 90% of Japan’s transportation in terms of tonnage. At present, logistics companies are facing an increase in fuel and labor costs as well as competition for transportation fees, and efficiency in logistics is required. The Ministry of Land, Infrastructure, Transport and Tourism (MLIT) built a logistics support system that transmits ETC2.0 probe data to logistics companies. In this study, we conducted a field experiment to examine the effectiveness of the support service for vehicle logistics management using ETC 2.0 probe data, which uses logistics support system. As a result, we were able to verify that the system is particularly valuable for logistics companies when transporting certain route. On the other hand, the problem became clear that there is a possibility of missing probe data and collection delay on regular roads depending on the route. We confirmed that these issues are solved by the use of “the roadside unit at the logistics center” and prepared functional requirements for their actual operation.

Abstract

Most QA systems are oriented to non-professional fields, but few of them are applied and analyzed in public transport and other professional fields. In this paper, we analyses the public transport question and answer corpus data, classify the text data by LDA topic model,...Lead moreMost QA systems are oriented to non-professional fields, but few of them are applied and analyzed in public transport and other professional fields. In this paper, we analyses the public transport question and answer corpus data, classify the text data by LDA topic model, then apply knowledge graph and information extraction technology to help reasoning to produce the appropriate answers. It can save time for the public to get answers, reduce the manual labor of urban public transport management departments, and assist the government in providing a decision support scheme and reasonable planning public transport network.

Abstract

The Metropolitan Expressway that provides arterial transportation in Japan’s Tokyo Metropolitan Area has a total length of 320.1 km, 14% or 43.2 km of which are tunnel structures. The Yamate Tunnel, which opened completely to traffic in March 2015, is a large, 18.2 km-lon...Lead moreThe Metropolitan Expressway that provides arterial transportation in Japan’s Tokyo Metropolitan Area has a total length of 320.1 km, 14% or 43.2 km of which are tunnel structures. The Yamate Tunnel, which opened completely to traffic in March 2015, is a large, 18.2 km-long tunnel with a daily two-way traffic volume of over 90,000 vehicles. Since the opening of the tunnel’s first segment in 2008, the Metropolitan Expressway Company has been implementing comprehensive disaster/safety management systems, including the most advanced disaster management equipment, a central control room system, optimized operations during fires, coordination with the fire department and other relevant agencies, and publicity/education for road users. This paper reviews the actual fire incidents that occurred during the past 10 years and discusses the measures implemented based on the lessons learned from them, as well as remaining challenges.

Abstract

Real-time network control strategies such as congestion tolling are widely used for traffic congestion mit- igation in urban transportation networks. Advances in Global Satelite Navigation System technology have led to increasing interest in distance- or usage-based tolli...Lead moreReal-time network control strategies such as congestion tolling are widely used for traffic congestion mit- igation in urban transportation networks. Advances in Global Satelite Navigation System technology have led to increasing interest in distance- or usage-based tolling as an effective alternative to traditional facility-, cordon- and area-based tolling that typically rely on fixed infrastructure. We propose a framework for pre- dictive distance-based toll optimization under variable demand. Given a set of charging zones, tolling function parameters are optimized using a simulation-based Dynamic Traffic Assignment (DTA) model operating within a rolling horizon scheme. Predictive optimization is integrated with the guidance information generation. Be- havioral models capture drivers’ responses to the tolls in terms of trip cancellation, mode, route and departure time. Evaluation of the proposed framework and solution algorithm, using the real-world Expressway network of Singapore, demonstrates the effectiveness of distance-based toll optimization in improving total social wel- fare compared to a traditional tolling scheme.

Abstract

Interest in Autonomous Vehicles (AVs) is gearing up tremendously in recent years with a focus on improving the safety and efficiency of the transportation system around the globe. Singapore being on the forefront of the deployment of AV related technologies for daily tran...Lead moreInterest in Autonomous Vehicles (AVs) is gearing up tremendously in recent years with a focus on improving the safety and efficiency of the transportation system around the globe. Singapore being on the forefront of the deployment of AV related technologies for daily transportation needs, is playing a pivotal role in testing and validation of AVs in urban road traffic scenarios. Several industries and research organizations are collaborating in Singapore for bringing AVs on road. This work presents the development of an integrated realistic simulator that enables different sectors involving automotive, communication, infrastructure, instrumentation etc., to analyse their developed modules for deploying AVs. The integrated simulator is unique in its way that it includes physics-based customizable models for sensors, facility to test different climate conditions and usage of standardized open formats. In order to test the simulator performance, a simple use-case is analysed for path planning and control for autonomous valet parking. The aim of this platform is to be used for independent tests and analysing how AV will behave in the urban Singapore like scenarios. This will help in deploying AV services for various applications in different regions in Singapore.

Abstract

The National Institute for Land and Infrastructure Management (NILIM) has been conducting public-private research with 29 private companies since January 2018 to realize next-generation cooperative ITS (C-ITS) by combining a next generation of vehicles with infrastructure...Lead moreThe National Institute for Land and Infrastructure Management (NILIM) has been conducting public-private research with 29 private companies since January 2018 to realize next-generation cooperative ITS (C-ITS) by combining a next generation of vehicles with infrastructure (V2I/I2V), and it is scheduled for completion in March 2019. This research aims both to make road control efficient by allowing road administrators to use various information collected through vehicles and to provide road users with information on road and traffic conditions based on information collected from specific vehicles. This is an interim report on research results concerning the provision of look-ahead information (LAI) and the advancement of road control, which are under study in this research.

Abstract

In a connected vehicle network, vehicles can report their state information periodically, such as position and speed, so that other vehicles or the infrastructure can make informed decisions based on the probe reports. Inaccuracy in the reported probes can cause the syste...Lead moreIn a connected vehicle network, vehicles can report their state information periodically, such as position and speed, so that other vehicles or the infrastructure can make informed decisions based on the probe reports. Inaccuracy in the reported probes can cause the system to mis-estimate the state of the traffic. A snapshot of the traffic state around a signalized intersection can be taken to predict the vehicles’ arrivals at the intersection so that an optimal traffic signal switching plan can be computed. An accurate assessment of traffic state is essential and thus it is important to study effects of probe imperfections on prediction accuracy. In this paper, we study the effect of probe imperfections for both green and green-to-red phases, with different Global Navigation Satellite System (GNSS) positioning errors and different probe reporting intervals. The car movement model used in the prediction is calibrated using a Bayesian optimization approach so that the results can reflect the best prediction accuracy achievable in various scenarios. The experiments were conducted on a simulation platform with the vehicle movement model attempting to make accurate predictions. The predictions were then compared to the simulation to assess the prediction errors under different imperfection scenarios.

Abstract

In single lane type toll road, toll collection system detects vehicle in order to make an originating point of pricing control process and vehicle detectors are required to have high reliability. Our company has been using a conventional thru-beam type sensor. However, th...Lead moreIn single lane type toll road, toll collection system detects vehicle in order to make an originating point of pricing control process and vehicle detectors are required to have high reliability. Our company has been using a conventional thru-beam type sensor. However, the detector arose some issues which were not able to solve because of the detecting principle of the thru-beam type light sensor. Therefore, our company has been developing laser scanning type vehicle detector. Yet, specific issues also exist in the laser scanning type. Our development aims to solve the issues of laser scanning type and also to obtain reliability comparable to beam type. In this paper, we describe our approach toward achievement of these targets.

Abstract

This study provides a new perspective of direct management of the air pollution in terms of urban road traffic. First, an estimation model for the level of area-wide air pollution is studied. Such study is conducted with the properties of the macroscopic fundamental diagr...Lead moreThis study provides a new perspective of direct management of the air pollution in terms of urban road traffic. First, an estimation model for the level of area-wide air pollution is studied. Such study is conducted with the properties of the macroscopic fundamental diagram (MFD), which represents the aggregated traffic state of a specified urban region. The model is confirmed by comparing the simulated and real-world data on the macroscopic traffic property. Second, based on the estimation model, the effects of different network perimeter control policies on air quality of urban region are tested with simulation experiments. The tests show that the perimeter control policies give positive or negative impacts on the area-wide air pollution depending on the traffic demand patterns. This study shows a possibility of new development of signal control strategies, which can be established upon the analysis on the influence of the direct traffic control on the air pollutant emissions in the urban area.

Abstract

Recently, according to the 4th Industrial Revolution, autonomous vehicles have been developed and many people are doing research and developing many new technologies related to them. One of them is Korea’s C-ITS(Cooperative-ITS) which is in operation to realize the import...Lead moreRecently, according to the 4th Industrial Revolution, autonomous vehicles have been developed and many people are doing research and developing many new technologies related to them. One of them is Korea’s C-ITS(Cooperative-ITS) which is in operation to realize the important V2V, V2I, V2X in autonomous driving. In addition, as the latest technologies such as IoT, Big Data, and Cloud technology have developed, the concept of Digital Twin has arrived. So, this study deals with the efficiency, safety, and environmental performance of the C-ITS when Digital Twin is used. Plus, mobile phone. It also discusses what if we can easily access the virtual world of a digital twin with your mobile phone and get information that you generate from C-ITS. It is the purpose of this paper to try to figure out how to get the maximum benefit in C-ITS from using Digital Twin.

Abstract

Sensor malfunction is the main reason of autonomous vehicle accident. This allowed authors to focus on road network space reservation control. Previous Studies focused on the situation judgement of autonomous vehicle rather than space reservation system and the concept of...Lead moreSensor malfunction is the main reason of autonomous vehicle accident. This allowed authors to focus on road network space reservation control. Previous Studies focused on the situation judgement of autonomous vehicle rather than space reservation system and the concept of network control was week. Our study has differentiated form other studies by considering about the foregoing elements. In this study. First, we analyse the necessary data and investigated the actual road data. We implemented it on VISSIM. Then, use MATLAB to data processing. The Network Space Reservation Algorithm process each car data through GIS approach by analysis of lane changing risk, autonomous car’s route, each lane’s traffic density, etc. By Network Space Reservation Algorithm, it is possible to increase safety, increase fuel efficiency, and increase the efficiency of the entire network. Also, there is a possibility of development of autonomous driving car road network control.

Abstract

This project investigates the implications of automated steering functions in current and potential near future model vehicles for road operations. The findings from this investigation support road agencies to optimise road design, maintenance and operation for automated ...Lead moreThis project investigates the implications of automated steering functions in current and potential near future model vehicles for road operations. The findings from this investigation support road agencies to optimise road design, maintenance and operation for automated vehicles with the primary aim of reducing road trauma on Australia and New Zealand’s roads. Findings generated through this work will better inform future changes to road authority pavement marking guidelines, and Australian Standards.

Abstract

The purpose of this study is to develop travel time prediction models that can take into account the spatial and temporal characteristics of the travel route. The recurrent neural network (RNN) model was employed to predict travel time. The consideration of the spatial-te...Lead moreThe purpose of this study is to develop travel time prediction models that can take into account the spatial and temporal characteristics of the travel route. The recurrent neural network (RNN) model was employed to predict travel time. The consideration of the spatial-temporal characteristics of this study is proposed to solve the time-lag problem. As results of the prediction, the 1 minutes ahead RNN model of Seoul to Kiheung was estimated to be the highest prediction accuracy. The MAPE and RMSE of 5 minutes ahead RNN model of Seoul to Kiheung were estimated to be 0.047 and 7.10 minutes, respectively. For the Seoul to Daejeon model, the 1 minutes ahead RNN model was estimated to be the highest prediction accuracy. The MAPE and RMSE of 1-minute ahead RNN model of Seoul to Daejeon were estimated to be 0.052 and 7.44 minutes, respectively.

Abstract

In this paper, a novel approach towards generation and implementation of local reference trajectory points is presented. The need arises because Autonomous Vehicles (AV) are intensively dependent on the absolute self-localization with respect to a global map for their pre...Lead moreIn this paper, a novel approach towards generation and implementation of local reference trajectory points is presented. The need arises because Autonomous Vehicles (AV) are intensively dependent on the absolute self-localization with respect to a global map for their precise motion. However, in several practical scenarios (such as heavy rain on a highway), localization with low covariance is nearly impossible. As a result, we propose an alternate approach to generate a reference trajectory using local cues such as curbs or lane markers, which can then easily be tracked by the AV using a suitable control algorithm. The key idea is to project the curb/lane points about a Frenet frame along the AV longitudinal direction. Further, minimization of computational load and implementation of a temporal filter for trajectory smoothing over time are also examined. Finally, the proposed theories are verified in experiments to provide smooth reference trajectory points.

Abstract

In Tokyo, the Rugby World Cup 2019 (RWC 2019) will be held in 2019 and the Tokyo 2020 Olympic and Paralympic Games (Tokyo 2020 Games) in 2020 are scheduled. At the Metropolitan Police Department(MPD), various transportation measures are being promoted towards the success ...Lead moreIn Tokyo, the Rugby World Cup 2019 (RWC 2019) will be held in 2019 and the Tokyo 2020 Olympic and Paralympic Games (Tokyo 2020 Games) in 2020 are scheduled. At the Metropolitan Police Department(MPD), various transportation measures are being promoted towards the success of the event, but as a countermeasure focusing on the audience who are coming to the competition site, a traffic signal that is a spectator route connecting the competition venue and the nearest station. In order to achieve smooth movement, we developed a new signal control, event-responsive pedestrian adaptive control, that we will introduce here.

Abstract

Recent trends in Autonomous Vehicle (AV) development has leveraged uses of low-cost light detection and ranging (LiDAR) sensors. While functionalities for enabling AV deployment have been of interest recently, however there are variant of mounting positions of the LiDARs ...Lead moreRecent trends in Autonomous Vehicle (AV) development has leveraged uses of low-cost light detection and ranging (LiDAR) sensors. While functionalities for enabling AV deployment have been of interest recently, however there are variant of mounting positions of the LiDARs with respect to the vehicle coordinate system. Furthermore, the number of LiDAR sensors implemented in an AV is also varying. This paper focus on solving technical issue when dealing with imprecise LiDAR mounting position by means of automatic-self-calibration. In addition, multi-LiDAR approaches requires accurate synchronization especially when AV is moving. It is shown that these solutions can correct discrepancies in pose and time errors and is independent of any methods to utilize LiDARs, e.g. for localization or perception purposes.

Abstract

East Japan Railway Company (JR-EAST) is the largest railway company in Japan, however, we also do real estate business such as development of station stores. When we investigate stations, it is important to know the information of station user, for example, time zone, sex...Lead moreEast Japan Railway Company (JR-EAST) is the largest railway company in Japan, however, we also do real estate business such as development of station stores. When we investigate stations, it is important to know the information of station user, for example, time zone, sex, age, and entrance station. Therefore, we proposed Consumption Potential Indicator (CPI) at ITS 2017. This is a new index combining the station users, residents and visitors, and annual income information. In this research, we have improved CPI to be a realistic indicator by considering spills and expenditure. Moreover, we developed a method to predict the future of CPI. We developed a tool that can compare CPI and other related indicators between 258 stations in the capital city of Japan. With this tool, we can comprehensively compare various indicators between arbitrary stations. Also, we can evaluate each station with new indicators considering actual human movements, so, it is possible to analyse the stations at finer granularity such as by time zone, sex, age, and entrance station, by time period. Therefore, we believe we can lead to development of commercial areas depending on customer’s needs.

Abstract

Since 2014, the transportation supervision mode has significantly changed, with the introduction of a series of credit management policies such as《Social Credit System Planning 2014-2020》. Under these policy frameworks, the transportation supervision department can clas...Lead moreSince 2014, the transportation supervision mode has significantly changed, with the introduction of a series of credit management policies such as《Social Credit System Planning 2014-2020》. Under these policy frameworks, the transportation supervision department can classify and supervise the market entities according to their degree of untrustworthi-ness. Market participants who are seriously untrustworthy will be se-verely restricted in areas such as bidding and government fund support, and even cross-disciplinary joint punishment. This paper studies the ef-fects of credit transportation policy, through the differences-in-differences model and using panel data on traffic violations of provinces from 2014 to 2017. The empirical research finds that the enforcement intensity of credit transportation policies at the provincial level is significantly nega-tively correlated with transport violations. Credit transportation policies have a relatively obvious effect on governance.

Abstract

It is known that removing traffic congestion takes time because of remained vehicles on streets. Therefore, it is crucial to prevent occurrence of congestion. This study examined the model to predict traffic congestion as the area in general (for the central area of Kyoto...Lead moreIt is known that removing traffic congestion takes time because of remained vehicles on streets. Therefore, it is crucial to prevent occurrence of congestion. This study examined the model to predict traffic congestion as the area in general (for the central area of Kyoto city) utilising spot traffic and average speed data measured by vehicle detectors in Kyoto, the most popular tourist city in Japan. Based on the model proposed in the previous study, the prediction model was upgraded by increase in learning data and utilising factors which should be considered to improve prediction accuracy (e.g. calendar, weather, and events). As the result, the prediction horizons with practical level of accuracy were extended to 24 hours.

Abstract

The Metropolitan Police Department (MPD) has been using an intelligent traffic control method called MODERATO (Management by Origin-Destination Related Adaptation for Traffic Optimization), which automatically calculates cycle lengths and green-time splits depending on tr...Lead moreThe Metropolitan Police Department (MPD) has been using an intelligent traffic control method called MODERATO (Management by Origin-Destination Related Adaptation for Traffic Optimization), which automatically calculates cycle lengths and green-time splits depending on traffic demand, at relatively big intersections of arterial roads. While cycle lengths and splits are successfully operated with upper and lower limits configured to secure enough crossing time and minimize waiting time for pedestrians, there has been challenges in controlling traffic lights when faced with “partial traffic congestion” where traffic congestion is detected only on one leg of an intersection of a major road and minor road(s). Therefore, the MPD introduced the cycle length control (hereinafter, “partial congestion control”), a new traffic signal control method which can automatically extend the cycle length to the preconfigured upper limit (a new upper limit which adds α seconds to the current upper limit), when faced with partial congestion, and recalculate the lower limit of the split for non-congested traffic streams entering the intersection so that more green time can be allocated to the heavy traffic entering the intersection to further alleviate congestion.

Abstract

Nowadays, Chinese cities are experiencing more and more serious traffic pressure and trying to get a balance between the city development and the capacity of transportation system. Big data and other new technology makes it possible to solve these problems, such as traffi...Lead moreNowadays, Chinese cities are experiencing more and more serious traffic pressure and trying to get a balance between the city development and the capacity of transportation system. Big data and other new technology makes it possible to solve these problems, such as traffic congestion and accident，and also influence transportation developing strategy. Firstly, this research summarizes the characteristics and problems of China on applying new technologies to transportation. Secondly, the main part of this research analyses the influence of such new technology, especially autonomous vehicle, on future transport systems from several different aspects, including traffic organization, service mode in order to figure out how future transport systems should be developed. Moreover, this research takes Guangzhou, a first-tier city in southern China, as an example, analyzing its current status of transportation development. Based on all these above, some strategies are provided for transportation developing in Guangzhou.

Abstract

It is an essential task to promote communication between vehicles and their surroundings to avoid potential traffic accidents. With the development of lighting technologies, it is becoming possible for vehicles to display more detailed information to other road users in a...Lead moreIt is an essential task to promote communication between vehicles and their surroundings to avoid potential traffic accidents. With the development of lighting technologies, it is becoming possible for vehicles to display more detailed information to other road users in an easier way. An exterior lighting system was proposed to display warnings on the road surface to call for the attention of cyclists. By displaying warning signals which can easily be understood, the cyclists can be informed about the future movements and actions of vehicles, even when they are not paying attention to the vehicles. However, it is still necessary to analyze whether the proposed exterior lighting system is effective or not. This study, therefore, performed a field experiment to investigate the influences of the proposed system on the behaviors of cyclists. The results indicated that traffic safety might be improved by applying the proposed exterior lighting system.

Abstract

This study adopts LSTM (from Deep Learning) to construct a prediction mode which introducing the real-time road traffic condition data of each road-section from the existing highway ETC system as model input variables, to predict the highway bus arrival time of a departur...Lead moreThis study adopts LSTM (from Deep Learning) to construct a prediction mode which introducing the real-time road traffic condition data of each road-section from the existing highway ETC system as model input variables, to predict the highway bus arrival time of a departure bus, then publish the information to users through the dynamic information system. The highway bus running time of the median/long distance transportation is a very important message for travelers. However, the running time is affected by the complex time-series, related with the road conditions changing by time and space. In the past, traditional prediction methods cannot construct an accurate prediction model. Sometimes, it leads to big errors where the traffic flow changing drastically. Compared with traditional historical statistical methods, the prediction results of this model can reflect the real-time traffic flow conditions and accurately predict the arrival time, effectively reducing errors.

Abstract

Auckland Transport is trialing ‘Dynamic Lanes’ along Whangaparaoa Road to reduce traffic congestion during peak times, testing a valuable tool for addressing congestion across the Auckland region. This is the first time in New Zealand a flush median has been used as part ...Lead moreAuckland Transport is trialing ‘Dynamic Lanes’ along Whangaparaoa Road to reduce traffic congestion during peak times, testing a valuable tool for addressing congestion across the Auckland region. This is the first time in New Zealand a flush median has been used as part of Dynamic Lanes to manage traffic flow. The system increases road capacity and safety by rearranging the full road width during peak hours, using a combination of LED line-marking and traffic control gantries to inform drivers which lanes are available to them. AT’s Travel Demand Manager shares the successes and behind the scenes of the dynamic lanes trial: https://www.youtube.com/watch?v=0gl1b8nj65o#action=share This Paper will explain the massive success of the trial and work through the unique issues faced and overcome including comments from users such as: • “It “was fantastic” being able to drive straight through the system” • “My usual trip was much faster, and I am pleased I was able to drive at the speed limit”

Abstract

When a transport project is to be built, it is critical to know how it is going to operate and what it is trying to achieve. Many new roads have been built without an understanding of the traffic impacts that are going to occur outside the project boundaries. In a rush to...Lead moreWhen a transport project is to be built, it is critical to know how it is going to operate and what it is trying to achieve. Many new roads have been built without an understanding of the traffic impacts that are going to occur outside the project boundaries. In a rush to open the new piece of infrastructure the operators of the network are often forgotten. This Paper will use the Waterview Tunnel Project in Auckland to explain how three transport operators, a builder and two road controlling authorities used Design Led Thinking and ITS technologies to create a smooth opening and operation of New Zealand’s largest piece of infrastructure.

Abstract

With ageing Public Address Systems and reduced penetration of traditional Radio Rebroadcast (RRB) solutions, the effectiveness of traditional road user evacuation methods is declining. It was proposed that targeted Flash SMS messages to affected road users could be prove ...Lead moreWith ageing Public Address Systems and reduced penetration of traditional Radio Rebroadcast (RRB) solutions, the effectiveness of traditional road user evacuation methods is declining. It was proposed that targeted Flash SMS messages to affected road users could be prove effective in co-ordinating the evacuation during an emergency. A system was developed to identify vehicles entering / exiting a tunnel, and leveraging customer records maintain a list of mobile phone numbers for road users in the tunnel at any time. It was anticipated that system could be used to provide a proportion of road users with specific evacuation instructions – improving evacuation responses, and compensating for declining effectiveness of traditional communications methods (Public Address and Radio Rebroadcasts). This paper provides a summary of the associated Safety Case, its findings and the role that Targeted Emergency Flash Messaging takes in providing the applicable Safety Hazard Control in these situations.

Abstract

The symmetric intersection design is a newly developed concept to tackle conflicts between left-turning and through traffic movements, where the four arms of the intersection alternate between left-hand traffic and right-hand traffic clockwise. Though the symmetric inters...Lead moreThe symmetric intersection design is a newly developed concept to tackle conflicts between left-turning and through traffic movements, where the four arms of the intersection alternate between left-hand traffic and right-hand traffic clockwise. Though the symmetric intersection design helps to increase capacity by reducing the number of signal phases needed, its counterintuitive layout may cause severe confusions to vehicle drivers and make it hard to arrange the right of way for cross-street pedestrians. This paper goes in depth on the mechanism of symmetric intersection design and discusses the feasibility and benefits of applying this design to autonomous vehicles. Furthermore, an improved signal control method is proposed to accommodate pedestrian crossing by splitting the typical three-phase timing scheme of symmetric intersection into equivalent multi-phase timing scheme. An optimization model is established to generate optimal signal timing for both autonomous vehicles and cross-street pedestrians. Algorithms are developed to solve this nonlinear programming optimization. The proposed methods are evaluated by VISSIM simulations with field data collected in Chengdu, China. Results show that the proposed methods significantly reduce the average waiting time and delay of autonomous vehicles, and provide pedestrians with necessary crossing time.

Abstract

To achieve “smart city”, advanced intelligent traffic system has recently been implemented in Macau to alleviate the complicated traffic situation. It includes a floating car data (FCD) system that collects real-time in-situ traffic data for computation of various traffic...Lead moreTo achieve “smart city”, advanced intelligent traffic system has recently been implemented in Macau to alleviate the complicated traffic situation. It includes a floating car data (FCD) system that collects real-time in-situ traffic data for computation of various traffic related indices. However, the existing FCD system cannot directly provide traffic flow (i.e., number of vehicles passing the road during a specific time) that are more useful for traffic planning. The purpose of this study is to develop a system that can be used to estimate the traffic flow based on FCD. An artificial intelligence approach, namely extreme learning machine, is employed to approximate the speed-flow relationship, and the resulting estimation system can provide prediction intervals of the traffic flow. Evaluation is conducted using real-world data in Macau, and the results show that the proposed system can provide reliable traffic flow estimation, which can assist Macau government for better traffic planning.

Abstract

In the vehicle warming, warning information induces vehicles to avoid potential risks. Warning information is attributed to the perceptual information in the vehicle information procedure. According to the heterogeneity of perception mode, this paper classifies the vehicl...Lead moreIn the vehicle warming, warning information induces vehicles to avoid potential risks. Warning information is attributed to the perceptual information in the vehicle information procedure. According to the heterogeneity of perception mode, this paper classifies the vehicle information procedure. The cyber information procedure is selected to construct the cyber vehicle running safety-warning model. The information greedy algorithm is adopted to solve this model. This paper evaluates the advantages of the model in the information utility, the safety state and the traffic efficiency through simulation experiments. The results show that compared with vehicle-level information, cyber information has better performance in the information effectiveness and the vehicle running safety. However, there is no significant difference in the traffic efficiency. Therefore, the application of cyber information in WVRS can improve the performance of vehicle running safety.

Abstract

Gui’an New District in Guizhou Province is planned to be the most dynamic growth pole in the southwestern part of the People’s Republic of China. It will be built into a safe, orderly, accessible, low-carbon, sustainable, and smart city. Intelligent transport systems will...Lead moreGui’an New District in Guizhou Province is planned to be the most dynamic growth pole in the southwestern part of the People’s Republic of China. It will be built into a safe, orderly, accessible, low-carbon, sustainable, and smart city. Intelligent transport systems will accelerate better and faster development of surface transportation in Gui’an New District and provide a base for the deployment of smart city functions. It will also support the operation and maintenance of safe, efficient, and green travel for the public. The Asian Development Bank is supporting a proposed project that aims to develop an interconnected system to support use, regulation, operations, and maintenance of all surface modes of transport such as pedestrian, buses, taxis, shared mobility services, bicycles, passenger cars, and freight vehicles. The project will also provide information on multi-modal hub traffic, law enforcement management, road construction and maintenance, and real time traffic operations monitoring.

Abstract

We, DENSO TEN, have developed an algorithm for parking space detection using cameras. The algorithm runs on a circuit board embedded in on-vehicle devices and has characteristics of a phased process divided into three steps, "linear approximation of outlines of parking li...Lead moreWe, DENSO TEN, have developed an algorithm for parking space detection using cameras. The algorithm runs on a circuit board embedded in on-vehicle devices and has characteristics of a phased process divided into three steps, "linear approximation of outlines of parking lines,” "parking line detection," and “parking space detection,” in each of which data is filtered based on our original models, to remove factors that have an adverse effect on the parking space detection. By evaluating this system at various types of parking lots in cities, we confirmed that it can detect parking spaces more accurately than conventional techniques do.

Abstract

We started demonstration project that we control PHVs charging status bundling many PHVs as one of electrical power resource of the Virtual Power Plant that provides systematic coordination power according to oversupplied/undersupplied grid situation. In this demonstratio...Lead moreWe started demonstration project that we control PHVs charging status bundling many PHVs as one of electrical power resource of the Virtual Power Plant that provides systematic coordination power according to oversupplied/undersupplied grid situation. In this demonstration, the parent aggregator order to increase / decrease the electric power demand to the PHV resource aggregator that we operate. We developed a system that enable to increase/decrease the electric power demand by moving the charging period of multiple PHVs by remote control. The system realizes both the user’s convenience of the vehicle and the power management function by taking into account the user’s PHV usage. This paper describes PHV charging power management method as VPP.

Abstract

The cooperative driving support systems using vehicle-to-vehicle communication are developed to avoid the accidents of crossing collision, left turn collision and so on. It is necessary to judge the relative lane with other cars accurately in order to notify the alert to ...Lead moreThe cooperative driving support systems using vehicle-to-vehicle communication are developed to avoid the accidents of crossing collision, left turn collision and so on. It is necessary to judge the relative lane with other cars accurately in order to notify the alert to the driver. We proposed the accurately relative lane decision using path history, instead of using the map information. We evaluated the proposed method with the log collected by actual driving.

Abstract

Traffic control optimization is a challenging task for various traffic centres in the world and majority of approaches focus only on applying adaptive methods under normal (recurrent) traffic conditions. But optimizing the control plans when severe incidents occur still r...Lead moreTraffic control optimization is a challenging task for various traffic centres in the world and majority of approaches focus only on applying adaptive methods under normal (recurrent) traffic conditions. But optimizing the control plans when severe incidents occur still remains a hard topic to address, especially if a high number of lanes or entire intersections are affected. This paper aims at tackling this problem and presents a novel methodology for optimizing the traffic signal timings in signalized urban intersections, under non-recurrent traffic incidents. The approach relies on deploying genetic algorithms (GA) by considering the phase durations as decision variables and the objective function to minimize as the total travel time in the network. Firstly, we develop the GA algorithm on a signalized testbed network under recurrent traffic conditions, with the purpose of fine-tuning the algorithm for crossover, mutation, fitness calculation, and obtain the optimal phase durations. Secondly, we apply the optimal signal timings previously found under severe incidents affecting the traffic flow in the network but without any further optimization. Lastly, we further apply the GA optimization under incident conditions and show that our approach improved the total travel time by almost 40.76%.

Abstract

Vehicle kinematic data have shown great potential for identifying aggressive driving behaviour in an objective and quantitative manner. However, the relationship between each kinematic indicator itself and aggressive driving still remains unclear. To address the issue, th...Lead moreVehicle kinematic data have shown great potential for identifying aggressive driving behaviour in an objective and quantitative manner. However, the relationship between each kinematic indicator itself and aggressive driving still remains unclear. To address the issue, this paper conducts an exploratory analysis to identify the kinematic indicators that have significant associations with driving behaviour. Based on kinematic data taken from the on-board unit of an experimental vehicle, 63 kinematic features were tested against various driving scenarios for statistical significance. The results suggest that 9 features, i.e. lateral acceleration and deceleration jerk, minimum vertical acceleration, mean pitch velocity, maximum yaw velocity, yaw angular deceleration, overall acceleration and deceleration and rate of change of heading direction, are significantly associated with aggressive driving behaviour. This study provides a reference for future studies to explore further relationships between kinematic data captured in on-board units and driving behaviour.

Abstract

When the operation of the train is disturbed due to accidents or weather, the train dispatchers operate train arrangements to return to the original timetable. Despite skillful skill of the train dispatcher is necessary for train arrangement, it has not been studied that ...Lead moreWhen the operation of the train is disturbed due to accidents or weather, the train dispatchers operate train arrangements to return to the original timetable. Despite skillful skill of the train dispatcher is necessary for train arrangement, it has not been studied that what kind of train arrangement is good. In this research, we devised Live SCORE, which is an index that estimates the loss time of passengers and quantitatively evaluates train arrangement. We created this index based on information acquired in real time. Therefore, we are able to use it in real time for the evaluation of operation adjustment. As a result of verifying Live SCORE, we found this index is meaningful. We believe that these indexes like Live SCORE become necessary when train operation become automated in the future and train arrangement also become more automated.

Abstract

Conventional control methods in Japan determine control parameters (cycle length, split and offset) sequentially, but the issue of whether the parameters are globally optimal remains. In response to this, we had developed a control method that optimized simultaneously con...Lead moreConventional control methods in Japan determine control parameters (cycle length, split and offset) sequentially, but the issue of whether the parameters are globally optimal remains. In response to this, we had developed a control method that optimized simultaneously control parameters. Our method adopts meta-heuristics to calculate parameters and needs high computation capability. Therefore, it has not been deployed widely due to installation of additional servers. On the other hand, the advancement of AI technology in recent years has been remarkable. It is difficult to explain why some AI-based methods calculate results. Therefore, there are serious issues that lack of accountability regarding results prevents application of these to mission-critical areas. Then we proposed to emulate our meta-heuristics method with deep learning in order to solve high computation capability of it and lack of accountability in AI-based methods. Our proposal was verified with a traffic simulation experiment.

Abstract

As the rise of internet of things and automatic driving technology, intelligent and connected vehicles have become one of the effective ways to improve urban traffic and reduce traffic accidents in the future. However, the security attacks and privacy threats to the intel...Lead moreAs the rise of internet of things and automatic driving technology, intelligent and connected vehicles have become one of the effective ways to improve urban traffic and reduce traffic accidents in the future. However, the security attacks and privacy threats to the intelligent connected vehicles (ICV) are critical to driving safety and people’s lives, which means a trusted and reliable solution has an important significance to ICV. This paper introduces the security architecture of ICV and analyses the challenges. Then a unified network identity authentication mechanism for vehicle-to-everything (V2X) is proposed to achieve digital signature and anonymous certificate on infrastructure-to-vehicle (I2V) messages.