Archives Search

Abstract

Since probe data can indicate driving situations by speed and time stamped data, it can be used as an alternative method for fuel consumption estimation as consumption rate depended on driving situations, which locally fluctuate following traffic conditions. This study ai...Lead moreSince probe data can indicate driving situations by speed and time stamped data, it can be used as an alternative method for fuel consumption estimation as consumption rate depended on driving situations, which locally fluctuate following traffic conditions. This study aims to determine time sharing of driving modes based on probe data, estimate fuel consumption, and study the impacts of using hybrid vehicles in Bangkok Metropolitan and Region (BMR) that traffic conditions varies by sub-area, day, period, and road category. 10,000 probe vehicles operated over BMR was collected. Meanwhile, fuel wasted in each driving mode and speed range was collected by mobile measurement from tested hybrid cars. Moreover, crawling is considered as the additional driving mode in this study. As a result, time sharing that represented time friction of each driving mode can well reflect traffic congestion by indicating stopping and crawling modes. Inner area of Bangkok on weekdays, minor roads, and accelerating mode are the greatest consumption of each case. Fuel consumption increases by high stopping and crawling percentages caused by traffic congestion. Under the assumption that all probe vehicles be replaced by the hybrid ones, fuel reduction impacts are clearly observed in congestion periods.

Abstract

In pedestrian-vehicle communications, pedestrian positioning via GPS plays an important role in protecting pedestrians from being collided by vehicles. Its precision, however, is greatly degraded in urban areas, where satellite signals are often obstructed by roadside bui...Lead moreIn pedestrian-vehicle communications, pedestrian positioning via GPS plays an important role in protecting pedestrians from being collided by vehicles. Its precision, however, is greatly degraded in urban areas, where satellite signals are often obstructed by roadside buildings. In this article, we propose a pedestrian positioning method by combining signals from both GPS satellites and vehicles. Using vehicles as anchors helps to solve the problem of satellite shortage. To improve the distance accuracy, a pedestrian estimates the distance to each nearby vehicle based on the signal strength of the line-of-sight component instead of the non-reliable overall signal strength (RSSI). In addition, from all the available vehicles and satellites, some of them that lead to a small positioning residue are selected for actual positioning. Extensive simulation evaluations, via ray tracing, confirm that the proposed method greatly improves the positioning precision and can reduce the horizontal positioning error to 3.31m, even under the practical limit of time resolution. And the positioning error can be further reduced to 1.83m when a finer time resolution is available.

Abstract

Recently, there is a trend that more and more workers are willing to commute by bicycles in the metropolitan area in Japan, since the bicycle is considered to be a healthy and eco-friendly travel mode. However, residents living in the local city are reluctant to use it fo...Lead moreRecently, there is a trend that more and more workers are willing to commute by bicycles in the metropolitan area in Japan, since the bicycle is considered to be a healthy and eco-friendly travel mode. However, residents living in the local city are reluctant to use it for commuting, since the commuting distance is very long, and it will spend much more time compared to the private car. In this study, we introduced a social experiment aiming at encouraging a mode shift of commuting from private cars to bicycles in Toyota City, Japan. The smartphone application collected the activity data of 13 monitors for 3 months. The results of the experiment suggest: 1) that health is most important motivations for the participants who are conversion from private cars to bicycles; 2) that real-time information provided by Facebook affects motivation improvement of bicycle commuting.

Abstract

Variable speed limit (VSL) and ramp metering (RM) affect freeway traffic operations in different ways and, accordingly, result in different effects on system travel time. The primary objective of this study is to propose an analytical procedure to help determine which con...Lead moreVariable speed limit (VSL) and ramp metering (RM) affect freeway traffic operations in different ways and, accordingly, result in different effects on system travel time. The primary objective of this study is to propose an analytical procedure to help determine which control strategy should be selected given different freeway bottlenecks and traffic conditions. The bottlenecks considered included an isolated merge bottleneck, a merge bottleneck with a closely spaced upstream off-ramp, and a diverge bottleneck with a closely spaced upstream on-ramp. Two RM and VSL control strategies were considered, including the ALINEA/Q and feedback based VSL. The improvements in system efficiency by various control strategies were tested using modified cell transmission models. A coordinated control strategy that combined ALINEA/Q and VSL control was also proposed. The effects of VSL and RM control strategies on system travel time at different freeway bottleneck areas were compared to identify the applicable conditions of different control strategies.

Abstract

It is important to understand the characteristics of older drivers to prevent traffic accidents caused by them. This study aimed to identify the factors determining the driving behaviour of older drivers. We hypothesise that the plausible factors might be individual drivi...Lead moreIt is important to understand the characteristics of older drivers to prevent traffic accidents caused by them. This study aimed to identify the factors determining the driving behaviour of older drivers. We hypothesise that the plausible factors might be individual driving style as well as self-regulation behaviour and compensatory strategies while driving. We collected data on styles and compensatory strategies of driving using questionnaires and assessed the self-reported health status, driving habits, and driving-related abilities. Factors affecting driving styles and drivers’ compensatory strategies were investigated using a Bayesian network model. The model revealed that the health status of the driver, such as presence of health problems and use of medicine, and the Useful Field of View (UFOV) performance influenced driving styles and drivers’ compensatory strategies. This indicated that medication influenced driving habits. In conclusion, the driving behaviour of older drivers could be inferred from their health status and UFOV performance.

Abstract

Drivers of ECs (Electric Cars) generally want to arrive at destinations within a planned time and consumed energy; however, the planned time and energy consumption cannot always be maintained due to traffic conditions and over consumption of energy. Violations of the plan...Lead moreDrivers of ECs (Electric Cars) generally want to arrive at destinations within a planned time and consumed energy; however, the planned time and energy consumption cannot always be maintained due to traffic conditions and over consumption of energy. Violations of the planned energy are especially critical for electric cars because violations increase the possibility of running out of stored energy. Therefore, driving support systems that analyze reachability and suggest actions to drivers are required. We proposed a method of reachability analysis in this paper to achieve driving support systems for ECs.

Abstract

affic congestion precursory has so far been computed by acceleration sensor via in-vehicle smartphone integration. The precursory is obtained from one vehicle motion, detecting the shock wave of traffic congestion. In our framework, predictive models are generated by (sta...Lead moreaffic congestion precursory has so far been computed by acceleration sensor via in-vehicle smartphone integration. The precursory is obtained from one vehicle motion, detecting the shock wave of traffic congestion. In our framework, predictive models are generated by (statistical) machine learning technique using the precursory because of dealing with latent (hidden) variables and of selecting a predictive model. In practice, our scheme allows a model to start with predicting the 30 minutes before traffic congestion, in order to cope with the non-stationary condition of congested flow. Therefore, it would favourably make use of the decision making of route-guidance and vehicle control system. By using predictive models, model adaptation is also feasible with cross national traffic data aggregated from Japanese and Indonesian highway. Conclusively, proposed prediction scheme using predictive models is shown for labelling congested flow in either heavy or weak.

Abstract

In recent years, the problem of hyper-congestions exists on Shanghai urban expressway. Simulation is expected to play an important role in evaluating various countermeasures for alleviating such hyper-congestions. However, reproducibility of current simulators becomes low...Lead moreIn recent years, the problem of hyper-congestions exists on Shanghai urban expressway. Simulation is expected to play an important role in evaluating various countermeasures for alleviating such hyper-congestions. However, reproducibility of current simulators becomes low when processing a very high traffic demand like peak hour traffic in Shanghai urban expressway. In this study, a traffic simulator has been developed which enables reproduction of traffic flow on Shanghai urban expressway especially under hyper-congestion. Firstly, data preparations are made including network building, OD matrix estimation. Secondly, gridlock problem in the existing simulation is detected. Thirdly, several improvements are made such as estimating two-phase density-speed function, proposing state-updating algorithm and designing node formation of virtual link. Then simulation results are compared to the observed traffic volume measured by traffic counts. Finally it is found that the developed simulator can reproduce traffic flow including the hyper-congestion at an acceptable level on Shanghai urban expressway.

Abstract

Traffic state estimation methods have been studied to predict change of service level in short term. However, if the target route in expressway has any capacity bottleneck section such as sag, it is difficult to predict when breakdown in flow happens. There is room for mo...Lead moreTraffic state estimation methods have been studied to predict change of service level in short term. However, if the target route in expressway has any capacity bottleneck section such as sag, it is difficult to predict when breakdown in flow happens. There is room for modelling the traffic flow dynamics from free flow state to congested state. Although past studies regarded parameters of traffic flow model as time-invariant, it is probable that the parameters are time-varying. The variation may be derived from randomness (i.e., white noise) or influence by longitudinal alignment or both. This paper tries 1) to estimate car-following behavior model equipping time-varying parameters with particle filter at sag section in critical traffic flow state using data collected on a driving simulator (DS), 2) to discuss how traffic flow state changes are detected, the averaged parameters of different courses in the DS are compared, and 3) to suggest a future direction of modeling breakdown in flow at bottleneck section and an idea to detect a sign of the breakdown.

Abstract

Thanks to the recent development of the motorway network in Japan, drivers have been able to choose alternative routes. However, there is no technical standard on the contents and information capacity of graphic route information panels (GRIPs) that provides information s...Lead moreThanks to the recent development of the motorway network in Japan, drivers have been able to choose alternative routes. However, there is no technical standard on the contents and information capacity of graphic route information panels (GRIPs) that provides information such as the travel time, one of the contributing factors for route choice decision-making process. Therefore, some GRIPs providing excessive information are normally found. This paper discusses the appropriate information capacity of GRIPs. As a case study, we use a driving simulator on a motorway section with an existing GRIP installed 2km short of a junction. More specifically, we classified various information displayed on GRIPs into place-name information, travel time information, and network shape information, and calculated the coefficients to convert travel time information and network shape information into place-name information and examined the threshold of the information capacity that can be defined by the converted place-name information volume.

Abstract

The prediction of the city distribution in the central area is the important part for urban distribution nodes and channel setting, as well as the transportation resources configuration. This paper, based on the data of 27 blocks in Wuchang and Jianghan district, analyzes...Lead moreThe prediction of the city distribution in the central area is the important part for urban distribution nodes and channel setting, as well as the transportation resources configuration. This paper, based on the data of 27 blocks in Wuchang and Jianghan district, analyzes the factors influencing the distribution generation from the distribution mechanism. It extracts the significant factors through the Pearson correlation analysis. Finally, the t-s fuzzy inference forecasting model is established, by using the Gaussian mixture model to determine the fuzzy rule base of fuzzy inference model and former parameters, and the consequent parameters was determined by the least square method. Based on contrast test, it is found that both in training and forecasting, t-s fuzzy reasoning based on Gaussian mixture model prediction model, which has good stability and extensibility, is more accurate than multiple linear regression model.

Abstract

Carsharing offers its customers access to a fleet of shared vehicles in an on-demand fashion. Despite growing popularity, carsharing remains a niche product due to some inconveniences, i.e., from the customer’s perspective there is a limited availability of the vehicles a...Lead moreCarsharing offers its customers access to a fleet of shared vehicles in an on-demand fashion. Despite growing popularity, carsharing remains a niche product due to some inconveniences, i.e., from the customer’s perspective there is a limited availability of the vehicles and very often vehicles have to be booked in advance and returned to the pick up location; from the operator’s perspective a fleet of vehicles has to be appropriately sized and rebalanced to effectively meet the demand with supply. A potential extension of the traditional carsharing can be seen in Autonomous Mobility on Demand (AMOD), which provides a one-way carsharing with self-driving electric vehicles. The objective of this work is to provide framework for the operation of AMOD systems, specifically focusing on the empty vehicles’ rebalancing. This paper reports on a study that made use of SimMobility—an agent-based microscopic simulation platform—to model and evaluate different scenarios of AMOD systems. This framework allows us to take the traffic congestion into account and evaluate its effect on the system performance. The simulation results show that rebalancing of AMOD system leads to a decrease in customers’ wait times and in that sense, is more efficient than the system with no rebalancing.

Abstract

This research aims to predict the influence of introduction of micro-cars into a metropolitan traffic network, on traffic point of view. Firstly micro-cars are distributed in the traffic network together with conventional vehicles using an asymmetric multi-class stochasti...Lead moreThis research aims to predict the influence of introduction of micro-cars into a metropolitan traffic network, on traffic point of view. Firstly micro-cars are distributed in the traffic network together with conventional vehicles using an asymmetric multi-class stochastic traffic assignment model, in which asymmetric interactions between classes are considered. Then the total network travel time and total network travel distance are calculated for different micro-car ratios among total traffic, and these values for the with-micro-car network are compared with those for the without-micro-car network. The results suggest that the effect of introducing micro-car on the total travel time is slight if their speed can be allowed to be higher than 50 km/h.

Abstract

Traditionally, the identification process of accident hotspots has been conducted on the basis of number of accidents at a specific location or section and/or their severity. However, the accident dataset used for this process includes two limitations: not all of accident...Lead moreTraditionally, the identification process of accident hotspots has been conducted on the basis of number of accidents at a specific location or section and/or their severity. However, the accident dataset used for this process includes two limitations: not all of accidents were collected by the police (i.e., most minor accidents were not reported), and the traditional process uses the accident dataset recorded for the past two or three years (i.e., most data inevitably include time lag). To overcome these two limitations, this study proposes a new model for identifying hotspots based on citizen participation through social media. Since the proposed method uses the public as sensors of hazardous road information through on-line social media, individual’s minor accident experiences as well as latent risk factors can be collectable. To demonstrate this approach, the case study has been carried out for a small district in the city of Goyang, Korea where is one of the most popularized cities in social media technologies. Then, the proposed and traditional methods are simply assessed.

Abstract

Regular evaluation of the intersections’ level of service is required to monitor the operation of traffic lights and maintain the quality of urban traffic. However, conventional methods to measure the intersection control delays such as deploying floating cars or manual r...Lead moreRegular evaluation of the intersections’ level of service is required to monitor the operation of traffic lights and maintain the quality of urban traffic. However, conventional methods to measure the intersection control delays such as deploying floating cars or manual roadside surveys require large amount of manpower and monetary cost. This paper proposes a new method to evaluate the intersections’ level of service by using the smart card data from the public transport system. It has minor requirements on other traffic information which can be easily obtained. With a provided smart card database of individuals’ trip information, an intersection in Singapore is taken as an example to illustrate the evaluation procedure, including converting the individuals’ trip information to floating card data, and estimating the intersection control delay. Generally, this new method provides a feasible approach to estimate the intersections’ control delay and hence to evaluate the level of service. Compared to conventional survey methods, this approach provides results with similar quality, but has very low cost and better coverage in terms of time and space.

Abstract

Environmental impact has become an important issue since the serious disasters related to climate change have happened frequently worldwide in recent years. In Taiwan, the road sector has accounted over 90% CO2 emissions of transportation sector. One of the possible alter...Lead moreEnvironmental impact has become an important issue since the serious disasters related to climate change have happened frequently worldwide in recent years. In Taiwan, the road sector has accounted over 90% CO2 emissions of transportation sector. One of the possible alternatives to reduce CO2 is through the use of public transits especially demand responsive transit systems (DRTS). DRTS aims at providing efficient vehicle routes to satisfy customer requirements under limited resources. In order to reflect environmental cost in DRTS, a formulation of the multi-objective dial-a-ride problem with the consideration of eco-efficiency (fuel consumption and emission) is developed and a multi-objective evolutionary algorithm is constructed to find the Pareto optimal solution. Two objectives specifically considered in the research include customer disutility and environmental cost. The multi-objective evolutionary algorithm is solved through a revised genetic algorithm with non-dominated sorting techniques. Numerical experiments are conducted to justify the proposed approach and to evaluate the quality of Pareto solution.

Abstract

Dedicated bus lanes are usually set up on congested roads to ensure the bus quality of service, which will on the contrary reduce capacity and increase delay of the general traffic because of exclusively occupying one or more lanes by buses. Intermittent bus lane (IBL) is...Lead moreDedicated bus lanes are usually set up on congested roads to ensure the bus quality of service, which will on the contrary reduce capacity and increase delay of the general traffic because of exclusively occupying one or more lanes by buses. Intermittent bus lane (IBL) is an effective approach to solve this problem. This paper utilized the concept of IBL and proposed an intermittent bus lane for left-turn (IBLFL) method to increase the capacity of signalized intersections, in which the dedicated bus lane in the middle of the road can be intermittently used for general left-turn traffic. The IBLFL method considered both operation constraints of buses and general traffic by coordination among bus operation, intersection main signal and the pre-signal at IBLFL entering area. The optimization procedure was formulated as a Mixed-Integer-Linear-Programming (MILP) and can be solved by branch-and-bound method. The implementation method is also described in this paper including the design for the layout of IBLFL section and the variable signs. Test results drawn from time-space diagram and simulation evaluation show that the proposed IBLFL approach performed well in terms of increasing the capacity of intersection and reducing average delay for both transit and general traffic.

Abstract

Vehicle positioning is a crucial requirement of car navigation systems. Despite tremendous advances in GNSS technology, multipath of non-line-of-sight (NLOS) GNSS signals still is a major drawback, especially in urban canyon areas. Mitigating the multipath problem has bee...Lead moreVehicle positioning is a crucial requirement of car navigation systems. Despite tremendous advances in GNSS technology, multipath of non-line-of-sight (NLOS) GNSS signals still is a major drawback, especially in urban canyon areas. Mitigating the multipath problem has been addressed at the physical level (e.g., new antenna designs) and at the logical level (e.g., signal filtering). Improvement at the physical level requires extra hardware installation which may not be suitable for in-car navigation applications due to the compact size of navigation devices. The logical level is not practical in some cases; for instance, considering the signal filtering techniques, the environment around a car is very transient (especially when travelling in urban canyons) such that sometimes only NLOS signals reach the receiver, thus, signal filtering cannot be performed. As an alternative approach to detect NLOS signals, this paper proposes a technique for a cooperative vehicle environment, called navigation-to-navigation (Nav2Nav), in which machine learning algorithms and pseudorange corrections work together to detect NLOS signals. To evaluate the approach, various machine learning algorithms were experimented through simulations and the results indicated >90% accuracy in detecting a satellite set with/without a NLOS signal from all satellites sets available.

Abstract

Occlusion is a common and annoying problem in video surveillance system. To address this issue, this paper proposes a Fast Regions with Convolutional Neural Network (Fast R-CNN) based strategy to detect and classify driving vehicle in surveillance video. Based on rich fea...Lead moreOcclusion is a common and annoying problem in video surveillance system. To address this issue, this paper proposes a Fast Regions with Convolutional Neural Network (Fast R-CNN) based strategy to detect and classify driving vehicle in surveillance video. Based on rich features learned from Deep Convolutional Networks (DNNs), Fast R-CNN shows remarkable performance even the vehicle is truncated or occluded. Fast R-CNN includes a region proposals extraction module and a real detection and classification module for each region proposal. Except for the performance of vehicle detector, the quality and numbers of region proposals has significant effect on the final detection precision and system cost. Different from static images, the motion information of objects in video can be helpful to reduce the search scope of region proposals. So this study tries to refine the region proposals of the video frames by taking motion foreground image as additional cues for more efficiency. Experimental results illustrate that compared with the original Edge Boxes method both the recall of the region proposals and the vehicle detector’s final detection rate have been much improved. It indicates that this region proposals improvement approach is well suited for colour surveillance video data from stationary roadside cameras.

Abstract

In this work we investigate the advantages of decision fusion for convolutional neural networks (CNNs) in image vehicle classification. Two frameworks of decision fusion for CNNs are focused, including multi-results by multi-CNNs(MRMC) and multi-results by multi-layers of...Lead moreIn this work we investigate the advantages of decision fusion for convolutional neural networks (CNNs) in image vehicle classification. Two frameworks of decision fusion for CNNs are focused, including multi-results by multi-CNNs(MRMC) and multi-results by multi-layers of the same CNNs(MRML). Different layers or different CNNs may produce highly conflicting results, which makes decision fusion less sense. To improve the performance of decision fusion with conflicting evidences from different CNNs or different layers, an improved weighted combination approach by modifying the D-S fusion model is proposed, which combines the current fusion result and new evidence together with weights calculated by Jousselme evidence distances with each other. The experimental results show that, the proposed approach can improve the accuracy in both frameworks and outperform reference decision fusion methods with high conflicting evidences.

Abstract

This paper proposes a regional traffic control cloud implementation that is a cloud based design optimization with different ramp metering system (RMS) rates and avenues association control strategies based on the timing calculation of green light and waiting time by gene...Lead moreThis paper proposes a regional traffic control cloud implementation that is a cloud based design optimization with different ramp metering system (RMS) rates and avenues association control strategies based on the timing calculation of green light and waiting time by genetic algorithm (GA) or particle swarm optimization (PSO) method. It was applied at Hsin-Chu Science Park along national high way #1 intersected with 7 major avenues called from path 1 to path 7 at Zhubei area. Those major boulevards at road intersection have been constructed with vehicle detectors (VD), Closed Circuit Television (CCTV), Changeable Message Sign (CMS), Automatic Vehicle Identification (AVI), Intersection Traffic Control (ITC), RMS, GPRS, ADSL, and Application Peripheral (AP), etc. The system operational logic at regional traffic control is applied according to major real-time sensors with XML-type message. The specified road networks at different paths are controlled with three major different built-in associated cloud-based optimization laws. One is using smoothly partitioning-flow, cutting flow, and split-flow strategy, respectively. Two guided golden laws are obeyed as following. One is to let vehicle drivers with least minimized delay time at waiting and second is with maximum numbers of vehicles throughput for each traffic control.

Abstract

Learning from telematics or connected vehicle data is a key technology of Big Data analytics in the domain of intelligent transport systems. How to aggregate the continuous big data inputs collected by moving vehicles and generate insights, e.g., automatic detection of ma...Lead moreLearning from telematics or connected vehicle data is a key technology of Big Data analytics in the domain of intelligent transport systems. How to aggregate the continuous big data inputs collected by moving vehicles and generate insights, e.g., automatic detection of map updates, is not only an important research problem, but also a hot topic generating market values and impacts. In this paper, we study the problem of learning road events from streaming telematics data, and propose a novel online clustering based algorithm enabling automatic road event learning. The core algorithm is specially designed for handling streaming telematics data considering location, time, event content information, etc., which is also scalable for huge volume data streams. A case study on traffic sign learning from car camera sensor data verifies the performance of the proposed approach.

Abstract

Given a predefined multi-modal trip plan that a commuter intends to follow, we optimize the time to book the on-demand transportation services that occur in the trip sequence such that we minimize waiting times for all parties involved. In particular, the objective is to ...Lead moreGiven a predefined multi-modal trip plan that a commuter intends to follow, we optimize the time to book the on-demand transportation services that occur in the trip sequence such that we minimize waiting times for all parties involved. In particular, the objective is to minimize the waiting time of the commuter to get onto the private mode of transport after she has finished all the preceding legs of the trip as well as minimize the waiting time of the on-demand transportation service that will pick up this commuter. We propose several algorithms, describe the system architecture and demonstrate their efficacy through simulations.

Abstract

V2X (Vehicle-to-Any) communications including V2P (Vehicle-to-Pedestrian) communications has been drawing much attraction as well as V2V (Vehicle-to-Vehicle) communications. Crash warning application is one of an important and promising application of V2P communications. ...Lead moreV2X (Vehicle-to-Any) communications including V2P (Vehicle-to-Pedestrian) communications has been drawing much attraction as well as V2V (Vehicle-to-Vehicle) communications. Crash warning application is one of an important and promising application of V2P communications. Although necessary protocols such as DSRC (Dedicated Short Range Communication) and IEEE802.11p are ready to be equipped, communication quality has been not well discussed for V2P communications. In this paper, quality of services in V2P communications is evaluated and compared with that in V2V communications. In order to improve the quality, node clustering methods are proposed. The proposed methods can mitigate network congestion in V2P communications in which the number of nodes is likely to be more than that in V2V communications. Simulation results shows that the proposed methods increase 20% of the number of nodes that are accommodated to meet the crash warning application requirements.

Abstract

The purpose of this analysis is to evaluate the performance of RSE for DSRC in Gyeonggi-Do, Korea including RSE that was first installed on a trial basis in 2009 and officially installed from 2012 to collect traffic speed information on national highways in Korea. The rou...Lead moreThe purpose of this analysis is to evaluate the performance of RSE for DSRC in Gyeonggi-Do, Korea including RSE that was first installed on a trial basis in 2009 and officially installed from 2012 to collect traffic speed information on national highways in Korea. The routes for and targets of evaluation test include 27 units of RSE on National Highway No.38, 26 units of RSE on National Highway No.42 and 6 units of RSE on National Highway No.77. According to the result of evaluation test, in the accuracy of communication, 57 units showed an accuracy of communication more than 95% which is the pass criterion. In the accuracy of sectional detection, 57 units of RSE passed a test with a good results of an accuracy of at least 97.5%. In the accuracy of information provision, 57 units of RSE passed the criterion by showing an accuracy more than 91.7%, and in the adequacy of detection zone, 57 units of RSE passed the criterion by showing a success of communication within 50m.

Abstract

In logistics park, the functional areas layout and the traffic organization influence on each other, in this paper, the author studies the layout method of the functional areas combining traffic organization so as to make the layout more practical. Firstly, the model for ...Lead moreIn logistics park, the functional areas layout and the traffic organization influence on each other, in this paper, the author studies the layout method of the functional areas combining traffic organization so as to make the layout more practical. Firstly, the model for functional areas layout design is established, which is based on traffic organization. Secondly, the solution algorithm is proposed, which includes three parts. The first part is fuzzy clustering analysis which is used to cluster functional areas so as to determine the traffic network factors. The second part is genetic algorithm for the functional areas layout problem, the author uses three sections of codes to construct the chromosome encoding, which are area code, cutting code and cutting sequence code. The third part is traffic organization, compared with genetic algorithm and CPLEX, in order to accelerate the calculation speed, the author selects shortest path distance for calculation. And CPLEX would be used to calculate the actual distance after the final layout is obtained. Finally, the model and algorithm are applied in the Lianyungang Port Logistics Park to verify the feasibility. Compared with the base layout, the total truck distance can be reduced by 13.8% at most.

Abstract

In cold regions, spreading of deicing chemicals (spreading operation) is commonly conducted to improve safety by increasing friction on icy roads, and trained, experienced operators are essential for the successful implementation of spreading operations. To deal with the ...Lead moreIn cold regions, spreading of deicing chemicals (spreading operation) is commonly conducted to improve safety by increasing friction on icy roads, and trained, experienced operators are essential for the successful implementation of spreading operations. To deal with the aging of the maintenance workforce and the shrinking of the pool of experienced workers, we need to further utilize ITS technologies to improve spreading operations conducted by novice or other inexperienced workers. This study investigated the eye glance behavior of winter maintenance personnel during the spreading operation to understand how spreading operation by experienced workers differs from that by novice/other inexperienced workers, and how assistive technology affects spreading operation by experienced workers relative to that by novice/other inexperienced workers. The study found that inexperienced workers can raise their performance during spreading operations to a level equivalent to that of experienced workers by being provided with information on the surface conditions ahead. The experienced workers showed the same tendency as the inexperienced workers did. However, since some of them were thought to be adversely affected by the information device, the development and delivery of training focusing on ITS technologies for existing maintenance personnel will be required.

Abstract

Interest in freeway, the negative interactions between slow-moving trucks and other vehicles are a primary cause of hazardous conditions, which lead to crashes with larger speed variations. To improve operational efficiency and safety, providing a climbing lane that separ...Lead moreInterest in freeway, the negative interactions between slow-moving trucks and other vehicles are a primary cause of hazardous conditions, which lead to crashes with larger speed variations. To improve operational efficiency and safety, providing a climbing lane that separates slow-moving trucks from higher performance vehicles is frequently considered when upgrading geometrics. This study developed an operations strategy for freeway climbing lanes based on traffic conditions in real time. To consider traffic safety when designing a dynamic strategy to determine whether a climbing lane is closed or open, various factors, including the level of service (LOS) and the percentage of trucks, are investigated through microscopic simulations. A microscopic traffic simulator, VISSIM, was used to simulate freeway traffic streams and collect vehicle-maneuvering data. Additionally, an external application program interface, VISSIM’s COM-interface, was used to implement the proposed climbing lane operations strategies. Surrogate safety measures (SSM), including the frequency of rear-end conflicts and Delta_S, were used to quantitatively evaluate the traffic safety using an analysis of individual vehicle trajectories obtained from VISSIM simulations with various operations scenarios. It is expected that the proposed algorithm can be the backbone for operating the climbing lane in real time for safer truck traffic management.

Abstract

The number of hypothesis nodes in multi-hypothesis algorithm grows exponentially with time, which costs lots of computation time and memory. Several methods are proposed to reduce the number of hypotheses nodes by improving the generation method of hypothesis node, prunin...Lead moreThe number of hypothesis nodes in multi-hypothesis algorithm grows exponentially with time, which costs lots of computation time and memory. Several methods are proposed to reduce the number of hypotheses nodes by improving the generation method of hypothesis node, pruning the branches of multi-hypothesis tree, eliminating and merging the redundant nodes. Field test results show that the methods greatly reduce the number of hypotheses, and save the computation time and memory of Multi-Hypothesis Map Matching algorithm.

Abstract

Alps diversity DSRC + GNSS antenna with BDA (Bi-Directional Amplifier) design was already presented at 22nd ITS World Congress in Bordeaux. Our current intention is to present evaluation results of the final product together with Alps OBU (On Board Unit), tested in real o...Lead moreAlps diversity DSRC + GNSS antenna with BDA (Bi-Directional Amplifier) design was already presented at 22nd ITS World Congress in Bordeaux. Our current intention is to present evaluation results of the final product together with Alps OBU (On Board Unit), tested in real operating conditions. Diversity DSRC + GNSS antenna with BDA and OBU represents a complete solution for V2V (Vehicle-to-Vehicle) scenario. Together with sample applications made by Alps Electric, the system can be used as a perfect starting platform for developing, debugging and testing customer’s applications.

Abstract

This paper describes the application of COM technology in the software platform development of metro ISCS, using the GE (General Electric Co.） IP configuration tool named as Cimplicity HMI. Compared with traditional application development methods, the COM technology de...Lead moreThis paper describes the application of COM technology in the software platform development of metro ISCS, using the GE (General Electric Co.） IP configuration tool named as Cimplicity HMI. Compared with traditional application development methods, the COM technology demonstrates many desired advantages in Cimplicity HMI environment, including higher software configuration efficient, easy code management and good maintainability. Keywords：COM, GE Cimplicity，HMI Configuration, ISCS, Metro system.

Abstract

Much has been made of the impact of GPS based telematics over the last 15 years and typically the sector has focused on the technical capability of the technology. In this paper Messrs Graham and Ferguson will focus on the application of the data set to analyse identify a...Lead moreMuch has been made of the impact of GPS based telematics over the last 15 years and typically the sector has focused on the technical capability of the technology. In this paper Messrs Graham and Ferguson will focus on the application of the data set to analyse identify and potentially solve a range of issues related to driver safety, risk management and risk pricing for a lower road toll. Demonstrating the connection between real world driver behaviour using inertial based GPS data, personality surveys and cognitive assessments, Graham and Ferguson will show how it is now possible to combine the data to generate an index that becomes the benchmark for future evaluations. This approach helps organisations proactively reduce crash risk by identify-ing employees who require specific driver coaching and/or training. Over time it has profound implications for Insurance, Risk Management and Pricing and Workplace Safety.

Abstract

Electric vehicles witnessed huge growth in the past 2 years. As the city selling the most EVs in China, Shanghai has become a big cake that all EV companies want to share, making it the highlight of EV market study for the whole country. The study of Shanghai EV parc fore...Lead moreElectric vehicles witnessed huge growth in the past 2 years. As the city selling the most EVs in China, Shanghai has become a big cake that all EV companies want to share, making it the highlight of EV market study for the whole country. The study of Shanghai EV parc forecasting is important to how the city develops this industry and plans charging infrastructure. Shanghai’s EV parc forecasting and the analysis of the features of this city’s EV market will be discussed in the report.

Abstract

In our connected world a growing demand for fast, reliable and cost-effective traffic solutions is fuelling the need for intelligent transportation. While transport authorities have traditionally utilised various technologies to capture traffic information, including manu...Lead moreIn our connected world a growing demand for fast, reliable and cost-effective traffic solutions is fuelling the need for intelligent transportation. While transport authorities have traditionally utilised various technologies to capture traffic information, including manual surveys, detector loops, floating cars and Bluetooth sensors, to understand how their networks are used, one major source of data remains relatively untapped — real-time crowdsourced data. By leveraging information from transport users themselves, combined with their own authoritative data, transport authorities can now establish two-way communication through open data channels such as the ArcGIS Open Data platform.

Abstract

The IRF recognizes the need for a call to action to review and facilitate the deployment of Intelligent Transport Systems (ITS) in specific markets. When properly applied ITS has significant potential to be an effective means to increase Safety, Ease congestion, reduce po...Lead moreThe IRF recognizes the need for a call to action to review and facilitate the deployment of Intelligent Transport Systems (ITS) in specific markets. When properly applied ITS has significant potential to be an effective means to increase Safety, Ease congestion, reduce pollution, attract sustainable financing, and lead to economic development and political and social acceptance. To date, ITS are not widely distributed nor utilized thus the full potential of sustainable Mobility and Transport is not clear. The reason for this gap can be attributed to cultural differences, system integration, education, legislation and financing. First, we will conduct a SWOT analysis on countries where ITS have been successfully applied to uncover the main drivers to success in sustainable mobility. Second, we will conduct a SWOT analysis in an emerging country which currently experiences a high number of fatalities by road accidents, high congestion and pollution with the means or intent to invest in infrastructure. The results will provide decision makers with insight to quantify the benefits of optimal ITS implementation over time and its risks as to create the policies and strategies to achieve sustainable mobility.

Abstract

The city centres in Auckland are the most frequented origins and destinations in the region, generating a high number of passenger trips. Auckland does not generate sufficient passenger numbers using public transport to provide significant relief to the road network. Curr...Lead moreThe city centres in Auckland are the most frequented origins and destinations in the region, generating a high number of passenger trips. Auckland does not generate sufficient passenger numbers using public transport to provide significant relief to the road network. Currently there are around 47,000 more cars than in the previous year, an increase of 5.5 percent in one year. A new idea “semester ticket in public transport for all university and vocational students” has been proposed in this work to narrow the gap between actual and expected number of public transport users. The analysis shows that Auckland can save approximately 12.7 million litres of petrol per year. An environmental emission analysis shows a reduction of 29 million tons of CO2 and around 740 thousand tons of CO emissions. There would be at least 19,600 less vehicle trips per day or 1,570 less vehicle trip per hour in peak hours. This would be useful easing of congestion in Auckland. This simple idea, which assists in optimising the transport system by improving public transport and increasing future demand for public transport, sustainable development, will have great importance.

Abstract

Automatic Incident Detection (AID) can deliver improved response times to incidents that cause congestion, injuries or traveller delay, providing a better experience for road users. Infrastructure costs for effective AID can be high, and significant false positive rates t...Lead moreAutomatic Incident Detection (AID) can deliver improved response times to incidents that cause congestion, injuries or traveller delay, providing a better experience for road users. Infrastructure costs for effective AID can be high, and significant false positive rates tend to reduce the benefits, if true incident reports are not acted upon. Existing systems may use Freeway Data Station sites, radar or fixed CCTV cameras as data sources to detect incidents. This paper describes a trial implementation of AID using an existing ‘dark’ optical fibre core, installed along a freeway corridor, connected to a unit performing Distributed Acoustic Sensing to detect and analyse traffic movements and any other acoustic signals such as a pedestrian on the road reserve to identify reportable incidents.

Abstract

The tasks of network modelling in metropolitan landscapes have a well established history in exploring strategic population growth infrastructure planning. This has been lead by the traditional four stage demand modelling approach, with variants for tour based exploration...Lead moreThe tasks of network modelling in metropolitan landscapes have a well established history in exploring strategic population growth infrastructure planning. This has been lead by the traditional four stage demand modelling approach, with variants for tour based explorations. However few tasks have explored the details of the intersection controllers as the constraint to identify how each approach at a signalised intersection holds a unique throughput opportunity. Such capacity measures utilise big data concepts into the modelling space, to explore how planning in Smart Cities can be influenced by the engineering constraints. This paper explores development of a metropolitan model for Melbourne, Victoria with an increased focus on matters of the supply element, rather than limited to the demand components. The model developed can be utilised for explorations into operations and planning for short and medium terms, as well as instigation of special events. The network development utilises the specifics of the intersections as well as a number of components related to the SCATS outputs for calibration of behaviour to each location.

Abstract

Container transport is a major challenge. In the 2014-2015 financial year, over 7 million TEUs (twenty-foot equivalent units) were processed in Australia’s ports. The vast majority of these are transported by truck. Planning this transport is a complex problem. In this pa...Lead moreContainer transport is a major challenge. In the 2014-2015 financial year, over 7 million TEUs (twenty-foot equivalent units) were processed in Australia’s ports. The vast majority of these are transported by truck. Planning this transport is a complex problem. In this paper, we look at this problem from an optimisation perspective: how can we plan the truck movements required to transport containers from wharf to customer and back, so as to minimise the total cost. Cost in this case includes the cost of running vehicles (cost per hour, per km), demurrage charges for late delivery, and fees paid at the container terminal or yard. In this technical paper, we describe some of the features we developed for optimised routing of containers.

Abstract

In transport fleet optimisation, one plans a number of transport tasks such as collecting and delivering goods, making optimal usage of the available transport fleet. Such a fleet can be diverse and consist of rigid vehicles and articulated vehicles consisting of a prime ...Lead moreIn transport fleet optimisation, one plans a number of transport tasks such as collecting and delivering goods, making optimal usage of the available transport fleet. Such a fleet can be diverse and consist of rigid vehicles and articulated vehicles consisting of a prime mover (truck) and one or more trailers. In the latter case, trailers can often be disconnected from one truck and connected to another. Often, a transport company will have considerably more trailers than trucks, and trailers may differ in their features. In this paper, we look at the technical challenges in fleet optimisation of a diverse fleet including trucks and detachable trailers.