The automotive industry is currently facing an automated driving revolution. This technology is tightly linked to societal and economic challenges: minimisation of traffic accidents, fuel consumption, traffic congestion, parking demand, and providing mobility for an ageing population, as well as to customer needs toward more personalised services. This study aims at the analysis and presentation of the current state of the art and prospects of automated vehicles (AVs) from various perspectives. This study concentrates on revision of the critical technologies, estimation of the impact on social aspects, identification of legal issues, consideration of factors in commercial success through user acceptance, and foresight carried out by other researchers. The primary material was prepared by review and analysis of research papers, standards, regulations, roadmaps, and projects in the field of AVs technology and its implementation worldwide. A SWOT analysis was performed, and it was found that for rapid AV spread, technological solutions need to be made taking into account law and regulation; user acceptance and human–robot interaction need to be solved together as part of one system.

Connected vehicles and fully automated driving systems are the main objectives of the future transportation system. A safe interactive system that interacts with people and things is essential to achieve these objectives. In this context, a crowd intelligence system plays a key role in interactive system development. Crowd intelligence is a combined method of data collection, integration and analysis from devices such as the smartphones, wearables, vehicles and a wide range of Internet of Things applications to use them as sensors. This collective feedback-driven interactive method is opportunistic for the development of the future transportation system. In this study, a survey is conducted considering crowd-intelligence techniques for the transportation system. From this survey, various challenges of the intelligent transportation system have been outlined and crowd-intelligent solutions have been discussed. A layered structure of transportation system architecture is suggested considering various problems in each layer and its crowd-intelligent solutions. The crowd-intelligence-based mobility, traffic control, traffic prediction, parking solutions have been discussed in this survey. Moreover, the importance of crowd-intelligent techniques and its applicability is discussed for sustainable development of futuristic transport infrastructure.

Due to the fact that existing traffic flow forecasting methods cannot completely reflect the real time traffic situation of the road network, a new method of short-term traffic flow prediction is proposed based on deep learning in this study. Firstly, in order to improve the efficiency of data processing, a method of road network data compression is proposed based on correlation analysis and CX decomposition. Secondly, the traffic flow data are divided into trend term and random fluctuation term by spectral decomposition method, and the influence of trend term on prediction accuracy is removed. Finally, by combining a deep belief network model and a kernel extreme learning machine classifier as the prediction model, the essential characteristics of the traffic flow data are extracted by using DBN at the bottom of the network, and the extracted results are input into the kernel extreme learning machine to predict the traffic flow. The actual regional road network traffic flow data are tested to verify the effectiveness of the proposed short-time network traffic flow forecasting method. The results show that the proposed method can not only save 90% of the running time but also the average prediction accuracy of each road section can reach 92%.

‘Destination charging’ – in which drivers charge their battery electric vehicles (EVs) while parked at amenities such as supermarkets, shopping centres, gyms and cinemas – has the potential to accelerate EV uptake. This study presents a Monte Carlo-based method for the characterisation of EV destination charging at these locations based on smartphone users' anonymised positional data captured in the Google Maps Popular Times feature. Unlike the use of household and travel surveys, from which most academic works on the subject are based, these data represent individuals' actual movements rather than how they might recall or divulge them. Through a fleet EV charging approach proposed in this study, likely electrical demand profiles for EV destination charging at different amenities are presented. Use of the method is presented first for a generic characterisation of EV charging in the car parks of gyms, based on a sample of over 2000 gyms in around major UK cities, and second for a specific characterisation of hypothetical EV charging infrastructure installed at a large UK shopping centre to investigate the impact of varying the grid and converter capacity on the expected charging demand and level of service provision to the vehicles charging there.

The prediction of energy consumption is the primary goal of an intelligent energy management system (IEMS). Based on the actual road–traffic conditions, the vehicle energy consumption on the whole planned path can be predicted online by road condition recognition or speed sequence prediction. Because the speed sequence prediction required by the latter cannot accurately reflect the real dynamic characteristics of vehicle speed such as acceleration and deceleration changes due to the random factors of traffic or human beings, which will greatly affect the predicting accuracy, especially on the urban road with complex working conditions. Therefore, based on the analysis of the cumulative relationship between vehicle speed characteristics and energy consumption, this study proposes a prediction method of vehicle driving energy consumption based on the statistical characteristics of vehicle speed, regardless of the accuracy of the prediction of vehicle speed sequence, including the establishment of a long-term vehicle speed feature prediction model and energy consumption prediction model by BP and SVM algorithms. Finally, its rationality is validated based on the authentic data with an accuracy of about 95%, significantly improved compared with that based on long-term vehicle speed prediction.

As urban traffic congestion is on the increase worldwide, many cities are increasingly looking to inexpensive public transit options such as light rail that operate at street-level and require coordination with conventional traffic networks and signal control. A major concern in light rail installation is whether enough commuters will switch to it to offset the additional constraints it places on traffic signal control and the resulting decrease in conventional vehicle traffic capacity. In this study, the authors study this problem and ways to mitigate it through a novel model of optimised traffic signal control subject to light rail schedule constraints solved in a mixed-integer linear programming (MILP) framework. The authors’ key results show that while this MILP approach provides a novel way to optimise fixed-time control schedules subject to light rail constraints, it also enables a novel optimised adaptive signal control method that virtually nullifies the impact of the light rail presence, reducing average delay times in microsimulations by up to 58.7% versus optimal fixed-time control.

Nowadays the railway industry is beginning to give serious consideration to using intelligent traffic management systems (TMSs) in order to improve railway performance regarding train and passenger delays and robust use of capacity. The TMS is responsible for handling railway traffic once a disturbance happens. A fundamental input parameter of a TMS is the train positions, to be used for traffic re-planning purposes. Inaccuracy in the train positioning data could significantly influence the effectiveness of a TMS. In this study, the authors developed a framework to evaluate how inaccuracies in the train position reporting may affect the TMS performance. This is achieved by assessing the impact of adding inaccuracies to the train position reported to a simulated TMS as it handles operational disturbances in real-time. The performance of the TMS is analysed by considering variability in overall delay outcomes after re-planning based on using accurate/inaccurate positional data. They demonstrate the usefulness of their framework in determining the positional accuracy required for the effective application of a basic rescheduling system via an example on a bottleneck area. Results show how the positioning inaccuracies can affect TMS and thus the overall delay.

Public transit has a significant impact on minimising traffic congestion and reducing the cost of travelling in urban areas. It is necessary to evaluate the efficiency of the public transit operation in response to the individual traveller's demands for transit. This study aims to analyse the demand for transit with overlapping origin–destination (OD) pairs to enhance the efficiency of transit operations. To achieve this, disaggregated-level travel demand data, i.e. individual traveller's data are collected from an automatic fare collection system called smart card. The Kneedle algorithm is used to calculate the knee point of travel demand. The overlapping OD pairs, which are higher than the knee point value, are calculated and displayed in a map format. On the basis of the overlapping OD pairs, the demand-based overlap index for each bus route is defined to evaluate the efficiency of bus operations. The proposed method is applied to six districts with higher transit demands than other districts in Seoul. On the basis of the results, discussion on the action plans to enhance the efficiency of bus operations are presented. The method proposed in this study contributes to improving the efficiency of the bus system by reflecting individual users’ travel demands.

The modern multi-modal transportation system has revolutionised the landscape of public mobility in cities around the world, with bike-sharing as one of its vital components. One of the critical problems in persuading citizens to commute using the bike-sharing service is the uneven bikes distribution which leads to bike shortage in certain locations, especially during rush hours. This study offers a system, which provides predictions of both rental and return demand in real-time for each bike station by using only one model, which can be used to formulate load balancing strategies between stations. Five different architectures based on recurrent neural network are described and compared with four evaluation metrics: mean absolute percentage error, root mean squared logarithmic error, mean absolute error and root mean squared error. This system has been tested with New York Citi Bike dataset. The evaluation shows the authors’ proposed methods demonstrate satisfying results at both the global and station levels.

As one of the efficient travel demand management approaches, vehicle restriction policy has been widely implemented worldwide. However, non-compliant behaviours occur and undermine such a policy. Maintaining a long-term deterrent effect on individuals is crucial. This study conducts an empirical analysis to investigate whether the vehicle restriction policy can prevent non-compliant behaviours in the long run. On the basis of a newly implemented vehicle restriction policy, i.e. odd-and-even policy in Langfang, China, compliance duration that represents the period from the beginning of the policy to the first rule-breaking behaviour of an individual is investigated by applying survival analyses. Moreover, this study proves the existence of inertia in rule-breaking behaviour. Findings reveal that a considerable proportion of vehicles would commit their first offence within a short period of time after the policy coming into force. Inertia would further increase the frequency of rule-breaking behaviour. Thus, vehicle restriction policy can hardly maintain a long-term deterrent effect. This study provides implications for understanding the deterrent effect of transportation policies and offers insights into policy improvement.

With the advancement of location acquisition technologies, a large amount of raw global positioning system (GPS) trajectory data is produced by many moving devices. Learning transportation modes from the GPS trajectory data is an important problem in the domain of trajectory data mining. Traditional supervised learning-based approaches rely heavily on data preprocessing and feature engineering, which require domain expertise and are time consuming. The authors propose a deep learning-based convolutional long short term memory (LSTM) model for transportation mode learning, in which the convolution neural network is first used to extract deep high-level features and then LSTM is used to learn the sequential patterns in the data that uses both GPS and weather features, thus making the full use of spatiotemporal operations. The authors have also analysed the impact of the geospatial region on human mobility. Experiments conducted on the Microsoft Geolife data set fused with the weather data set show that their model achieves the state-of-the-art results. The authors compare the performance of their model with the benchmark models, which shows the superiority of their model having 3% improvement in accuracy using only GPS features, and the accuracy is further improved by 4 and 7% on including the impact of geospatial region and weather attributes, respectively.

Both steering assistance and differential braking have an irreplaceable advantage for man–machine cooperative control of lane departure prevention. To realise the human–machine cooperative control of lane departure prevention with minimal human–machine conflict, a multi-mode human–machine cooperative control algorithm based on steering assistance and differential braking is proposed. Taking the longitudinal velocity and road curvature as input and considering the driver torque, the fuzzy controller used to determine the virtual lane boundary is designed. The control system is divided into the driver control mode, the steering assistance cooperative control mode and the differential braking cooperative control mode by the driver state and the virtual lane boundary width. Next, the model predictive controller is developed to determine the front steering angle in the steering assistance cooperative mode and the additional yaw moment in the differential braking cooperative mode, respectively. Finally, the active disturbance rejection controller is used to determine the assistance torque, and single rear wheel brake control strategy is developed to calculate the wheel braking force. The effectiveness of the proposed multi-mode human–machine cooperative control algorithm is evaluated with numerical simulation and experiments on a human-in-the-loop platform.

This study considers the problem of vehicle platooning using multiple vehicle look ahead information while utilising a controller based on the constant spacing policy (CSP). First, string instability is shown with a CSP based controller that utilises position, velocity and acceleration information from ‘r’ predecessor vehicles in the presence of parasitic lags in actuation. A novel approach using perturbation analysis is employed to show this result, where the spacing errors are states of a spatially discrete system. Then, an upper bound on the allowable parasitic lag is derived when the CSP controller utilises information from the immediate predecessor vehicle and the leading vehicle; for this controller and a given parasitic lag, a design procedure for selection of the controller gain is also given. Numerical simulations of these two results are provided and discussed.

In-vehicle interfaces are now part of the vast majority of production vehicles. Such interfaces need to be thoroughly evaluated to ensure they do not pose any risks to the drivers using them. Driving simulators have extensively been used in such a context, yet their reliability in terms of how realistic a driving behaviour they elicit is still in question. An investigation on driving simulator behavioural validity in the context of prototype human–machine interface evaluation is presented in this study. Using data collected in a dual setting driving study (driving simulator and real world), as well as results from existing related literature, a comparison between driving behaviour in different types of driving simulators and in reality was carried out, for a variety of behavioural metrics. The results are presented in the form of a ‘validity matrix’ that aggregates the level of behavioural validity different simulator settings can achieve for different behavioural metrics.

Railway traffic conflicts are common in the day-to-day operation of trains due to the limited track capacity, the varying priority of trains, localised weather conditions, maintenance operations etc. The conventional conflict resolution strategy focuses on delaying the trains by considering the braking distance of a preceding train, cancellation or intermediate stopping of the trains etc. This strategy can solve the problem of railway operators than the passenger's problem of missed connecting trains, missed business opportunities or personal appointments etc. To ensure both operator and passenger satisfaction, this paper proposes a novel blockchain-enabled virtual coupling of automatic train operation fitted mainline trains for railway traffic conflicts. The immutable blockchain databases of the trains and track infrastructures help to forecast the traffic conflicts in real time. Seven variants of virtual coupling strategies are described in this study. Based on the chosen strategy, the reference model of the automatic train operation of mainline trains is virtually coupled or synchronised. Finally, the simulation results and theoretical analyses using several case studies are carried out to confirm the sufficiency of the proposed system and method. The major advantage of the proposed study is that it can be an overlay to the existing European Railway Traffic Management System Level-2.

Advanced driver assistance systems (ADAS) become an integral part of almost all modern automotive systems. ADAS have been evolving over a decade and the expansion of vision-based ADAS is quite rapid mainly due to the recent advancements in camera technologies. Most of the vision-based ADAS applications have been developed focusing on structured environment parameters and being tested adequately for those environments whereas they cannot be applied with their current framework as such for non-structured environments due to various limitations. This study presents a comprehensive overview of challenges in expanding the vision-based ADAS for non-structured environments. The authors have proposed a segmentation detection method for pedestrians and cyclists in a non-structured road environment to improve the accuracy of the popular deep learning networks. This method uses upper body detection and a pairing technique that improves the average precision significantly without consuming much computational resources. This approach would help to transform the structured environment ADAS to non-structured environments with minimal modifications. With the proposed approach, they are able to increase the accuracy of certain object classes up to 49% for various popular deep learning networks.

Speed sign is one of the most important instant indications for drivers to adjust the speed of their cars. In the literature, almost all of the existing methods for speed sign recognition are based on static pictures with clean images. When dealing with the traffic signs in complex environments, these existing approaches often have inaccurate detected areas, which lead to the difficulty of recognition. The authors propose a deep cascade network to improve the recognition of the speed signs with a structure of cascade subnetworks. The proposed network is composed of a localisation subnetwork and a classification subnetwork. The difficult issue in complex scenarios is the detection of the speed sign due to its small resolution, occlusion, colour fading etc. The proposed localisation subnetwork can improve the localisation accuracy by borrowing the idea of locating the targets from coarse to fine. Ultimately, the classification sub-network extracts more effective features for speed sign recognition. The experimental results illustrate that the proposed method outperforms the YOLOv2 or YOLOv3 model in identifying the speed sign in complex scenarios with at least 6% higher in terms of area under curve, and this will promote the improvement of recognition significantly.