The appraisal of intelligent transport systems (ITS) has become increasingly important in order to capture their full range of potential impacts. This study aims to assess the appropriateness of conventional transport appraisal models and tools for this task, particularly in reflecting the environmental and socio-economic impacts of ITS. These include the most common environmental systems analysis tools (ESATs), which incorporate international standards and are of considerable importance in indicating sustainability. A review of how emerging methods relate to the goal of a successful transition to a low carbon future is reported, based on the literature. The appraisal of ITS is inherently uncertain due to the decentralised nature of information communication technology; therefore, a range of methods to capture this aspect are reviewed. The models, weights and methods are analysed concerning their ability to estimate sustainability performance, given the numerous configurations of ubiquitous technology that may comprise ITS services. Weighting methods are important in reflecting perceptions of how sustainability should be assessed. These can be incorporated by identifying, classifying and selecting one or more ESAT's based on their suitability for a particular application. Finally, recommendations are given on which tools can be integrated to more comprehensively reflect the performance of ITS.

Automated lane detection is a vital part of driver assistance systems in intelligent vehicles. In this study, a multi-lane detection method based on omnidirectional images is presented to conquer the difficulties stemming from the limited view field of the rectilinear cameras. The contributions of this study are twofold. First, to extract the features of the lane markings under various illumination and road-surface scenarios, a feature extractor based on anisotropic steerable filter is proposed. Second, a parabola lane model is used to fit the straight as well as curved lanes. According to the parabola lane model, the straight lines and curves of feature maps can be represented as straight lines in a linear space coordinate system. Then lane modelling can be treated as an optimisation question in linear space and the parameters of lanes can be estimated by minimising the objection function. The method has been tested on publicly available data sets and the real car experiments. Experimental results show that the proposed method outperforms state-of-the-arts approaches and obtains a detection accuracy of 99% in real world scenes.

The existing tracking and recognition methods concentrate mainly on single-class targets; however, systems for traffic management or intelligent transport often require multi-class target tracking and recognition in real time. This study proposes an effective multi-class moving target recognition method that is based on Gaussian mixture part-based model, which accurately locates objects of interest and recognises their corresponding categories. The method is multi-threaded and combines soft clustering approach with multiple mixture part based models to provide stable multi-class target tracking and recognition in video sequences. The highlight of the method is its ability to recognise multi-class moving targets and to count their numbers in the video sequence captured by a stationary camera with fixed focal length. Another contribution of this study is that an extended part based model is developed for object recognition in real-world environments, which can improve the overall system performance, lower time costs, and better meet the actual demand of a video system. Experimental results show that the proposed method is viable in real-time multi-class moving target tracking and recognition.

The trajectory control of autonomous vehicles is an area which has attracted much research recently because it can prevent accidents caused by driver errors and significantly improve road capacity. Overtaking is one of the most complex and challenging manoeuvres for road vehicles and the autonomous control of the vehicle during this manoeuvre can greatly improve vehicle safety. As the innovative four-wheel independent steering (4WIS) and four-wheel independent driving (4WID) electric vehicle can provide redundant control actuators, this study focuses on utilising 4WIS–4WID techniques and vehicle dynamics control to achieve better control of autonomous vehicles. This study first introduces the traditional two-wheel proportional–integral–derivative (PID) steering control and two-wheel sliding mode controller (SMC) driving control for autonomous vehicle control. Then based on these, the four-wheel PID steering controller and four-wheel SMC steering controller are proposed. A four-wheel SMC driving controller and a four-wheel combined yaw rate and longitudinal velocity SMC driving controller are also proposed. Simulation results prove that the best control performance can be achieved when the four-wheel SMC steering controller and four-wheel combined yaw rate and longitudinal velocity SMC driving controller are used together.

The traffic safety on expressways is crucial for the efficient operation of the expressway system, and there is a close relationship between traffic states and crashes on expressways, and the occurrence of crashes may be influenced by the interaction of different combinations of traffic states upstream and downstream of the crash location. Based on the crash data and the corresponding traffic flow detector data collected on expressways in Shanghai, this study proposes a hybrid model combining a support vector machine (SVM) model with a k-means clustering algorithm to predict the likelihood of crashes. The random forest (RF) model is employed to select the important and significant variables for model construction from the data of the traffic flow 5–10 min before the crash occurred. Then, the cross-validation and transferability of different models (SVM model without variable selection, SVM model with variable selection, and hybrid SVM model with variable selection) are determined using 577 crashes and 5794 matched non-crash events. The results show that the crash prediction model along with the four most important variables selected using the RF model can obtain a satisfactory prediction performance for crashes. With the combination of the clustering algorithm and SVM model, the accuracy of the crash prediction model can be as high as 78.0%. Moreover, the results of the transferability of the three different models imply that the variable selection and clustering algorithm both have an advantage for crash prediction.

This study aims to find an experientially reliable path considering travel time uncertainty and driving experience of local probe vehicle drivers. Accordingly, a two-stage route-finding procedure is proposed. First, a set of candidate paths is built by using the hyperpath algorithm, where the choice probability is assigned to each link with uncertain travel time. Second, the shortest path algorithm is applied to find the experientially reliable path on the graph of hyperpath where the modified link cost is penalised based on the link choice probability derived from hyperpath algorithm and the driving experience of local drivers. Four kinds of optimal path in a real-world network are compared with the observed one. It is found that the proposed path has the most similarity with the observed path and it has a higher degree of familiarity and reasonable time and distance.

Emerging vehicular communication makes it possible to provide traffic light information to drivers inside vehicles with the application of in-vehicle devices. However, the effect of this method on driver behaviour is still unclear, and there is concern that the application of in-vehicle traffic lights may result in driver distraction. This study proposed two modes of in-vehicle traffic lights to assist drivers: a ‘current’ mode providing real-time information of the upcoming ground traffic lights, and a ‘predicted’ mode offering predicted information regarding ground traffic lights, taking into account the time to arrival at the upcoming intersection. Two kinds of in-vehicle devices were compared for displaying in-vehicle traffic lights: a normal 4.3-inch display and a head-up display. A driving simulator experiment was executed for 11 subjects, and driver behaviour was evaluated for driving operations and eye-gaze behaviour. The results demonstrated that disruptive braking and accelerating operations were significantly reduced under the predicted mode, and glance time was significantly shorter for the head-up display than for the normal 4.3-inch display. The authors concluded that the predicted mode easily prompts drivers to ecological driving, and that the head-up display is reliable for providing in-vehicle traffic light information.

The high variability of sign appearance with partial occlusions in uncontrolled environments has made the detection of traffic signs a challenging problem in computer vision. In this study, an occlusion-robust traffic sign detection framework is proposed. To achieve occlusion-robust detection, a colour cubic feature called colour cubic local binary pattern (CC-LBP) is proposed to construct a coarse-to-fine cascaded detector. The CC-LBP utilises colour information and a self-adaptive threshold to express multiclass traffic signs, which can effectively remove non-object subwindows in the cascade-based detection. The verification experiments show that the proposed CC-LBP feature performs better than the previous rectangular features in representing multiclass traffic signs, and that the proposed occlusion-robust detection method can detect multiclass partial occluded traffic signs with high accuracy in real time.

This paper demonstrates the effectiveness of video analysis for a cyclist’s data collection in high-density environments. It attempts to address the shortcomings of conventional data collection methods by conducting an automated study to obtain real-world bicycle data and providing a validation scheme to assess the accuracy of the automated observations. Basic traffic quantities such as average speed, volume count, flow rate, and density are automatically estimated and validated. Furthermore, traffic analysis applications are conducted on the collected data as a demonstration of the capabilities of the automated computer vision system. The analysis is applied to a data set collected through video cameras at a cycling event at the University of British Columbia. The analysis indicates the feasibility to automate the cyclist traffic data collection process in challenging, dense conditions. The reported results can provide a motivation for traffic engineers to rely on automated data collection as guidance during the decision-making process and to explore further the relationship between the bike facilities width, the expected flows, the facilities performance, and level of safety.

Electric vehicle (EV) charging problem impedes its wide scale commercial adoption. In this study, the authors address this problem through an ant colony optimisation based multiobjective routing algorithm that is dedicated to accommodate EV trips. By using connectivity, EVs communicate with other vehicles and infrastructure components to transmit information in real time for finding the best route, and for intelligently recharging on the move using an inductively coupled power transfer system. Such connected EVs are capable of adapting each trip with the lowest travel time and/or the lowest recharge cost along with an optimal recharge plan to prevent a battery drain. As a case study, a real world roadway network in Charleston, South Carolina was simulated to examine the performance of the routing strategy. Simulation analysis revealed that connected EVs can reduce not only the total travel time and the energy consumption, but also the recharged volume of electricity and corresponding cost, thus significantly relieving the concerns of range anxiety of EV drivers. This routing approach potentially leads to a reduction in the EV battery capacity requirement, which in turn can reduce the cost of energy storage systems to a reasonable level.