To address the current challenges of mobility in large urban areas, the authors wish to make practical the experience of multimodal travels, where a range of complementary transportation services are offered to an individual. Making multimodal travels is, however, still complex and the burden is let to the traveller. In this paper, the authors present their work for making the user experience of planning a trip in a multimodal mobility context simple and efficient. The authors recall the technical challenges and literature associated with such planning which must be truly multimodal, real-time and contextualised to the user preferences. The authors then describe the main features, differentiation, and advantages of the Xerox Trip Planner system, their solution for the computation of multimodal trips. The authors explain their path alternative generation method and a way to allow for more complex transfer mode sequences in the so-called round-based multimodal trip planning algorithms. The authors also present a comparative study with data from the city of Adelaide, Australia, which assesses the dynamic planning component of their solution with the other options available in this city.

Although rainfall affects travel speed on different road classes to different degrees, few existing studies focus on the impacts of rainfall on different urban road classes. Utilising real-time weather data, floating car data-based travel speed on urban roads and traffic performance index in Beijing, this study analyses and compares the differences between travel speeds for normal versus rainfall weather under different precipitation intensities, different congestion levels and different road classes. The study demonstrates that when precipitation intensity reaches the heavy rain level, travel speeds on the expressway, major arterial and collector decrease by 7.5, 5.0 and 9.4%, respectively, at night, and decrease by 15.2, 13.4 and 12.2%, respectively, during peak hours. Further, this study derives a relationship between precipitation intensity and travel speed reduction for different congestion levels. The light-congestion is the most sensitive traffic condition to rainfall when there is moderate or heavier rain. Finally, a correction model for travel speed predictions for rainfall weather is developed and applied. It is indicated that the model can predict travel speed under rainfall weather effectively.

For coastal surveillance, this study proposes a novel approach to identify moving vessels from radar images with the use of a generalised Bayesian inference technique, namely the evidential reasoning (ER) rule. First of all, the likelihood information about radar blips is obtained in terms of the velocity, direction, and shape attributes of the verified samples. Then, it is transformed to be multiple pieces of evidence, which are formulated as generalised belief distributions representing the probabilistic relationships between the blip's states of authenticity and the values of its attributes. Subsequently, the ER rule is used to combine these pieces of evidence, taking into account their corresponding reliabilities and weights. Furthermore, based on different objectives and verified samples, weight coefficients can be trained with a non-linear optimisation model. Finally, two field tests of identifying moving vessels from radar images have been conducted to validate the effectiveness and flexibility of the proposed approach.

Providing predictive information is expected to be an effective way in reducing congestion because travellers can have information of future traffic conditions and plan their travel ahead of time. However, due to the complex and computational expense of traffic flow forecasting, only short horizon future travel times can be generated and provided, which seems not sufficient for travellers to accurately plan their travel along an entire path. In this study, the authors have developed a simulation model to investigate the impact that limited time-span predictions have on the effectiveness of the predictive information. The results indicate that even though the predicted horizon is very short, the predictive information can still perform better than the current prevailing information. Sensitivity analysis of market penetration rates and prediction horizons is also discussed. This study is expected to be guidance for practical implementation and operation of short-term predictive information scenarios.

Ramp metering (RM) and variable speed limits (VSLs) affect freeway traffic in different ways and, accordingly, result in different effects on travel time and crash risk. This study compared the effects of RM and VSL at different freeway bottleneck segments with various traffic demands. An isolated merge bottleneck segment was examined, and then extended to consider closely spaced upstream ramps and downstream diverge bottleneck. Two control strategies were tested which were the Asservissement LINéaire d'Entrée Autoroutiére (ALINEA)/Q and feedback-based VSL. The reductions of travel time and crash risk were evaluated using the modified cell transmission models. The results showed that ALINEA/Q was more stable than the feedback VSL, but its power was limited by ramp space. A coordinated RM and VSL strategies were proposed to improve the control effects. The authors also evaluated how traffic demand features on mainline and ramps affected the control effects. The results highlighted the importance of consideration of geometrical features and traffic demands on freeways when selecting the best control strategy.

Driver drowsiness is a frequent cause of traffic accidents. Research on driver drowsiness detection methods is important to improve road traffic safety. Previous driving fatigue detection methods frequently extracted single features such as eye or mouth changes and trained shallow classifiers, which limit the generalisation capability of these methods. This study proposes a framework for recognising driver drowsiness expression by using facial dynamic fusion information and a deep belief network (DBN) to address the aforementioned problem. First, the landmarks and textures of the facial region are extracted from videos captured using a high-definition camera. Then, a DBN is built to classify facial drowsiness expressions. Finally, the authors’ method is tested on a driver drowsiness dataset, which includes different genders, ages, head poses and illuminations. Certain experiments are also carried out to investigate the effects of different facial subregions and temporal resolutions on the accuracy of driver fatigue recognition. Results demonstrate the validity of the proposed method, which has an average accuracy of 96.7%.

Urban arterials are characterised by high traffic volume, and driveway densities which cause congestion and crashes. In urban arterials, safety and operational issues can be improved by access management strategies. One such strategy is to restrict traffic entering the urban arterial to ‘right-in–right-out’ through implementing a raised median. While past research has shown the operational benefits of this strategy, it has not been evaluated in the context of dynamic access control. This study investigates the effectiveness of the connected vehicle (CV)-supported dynamic access control. The analysis is applied to an urban corridor under four scenarios: (i) the existing condition with direct left turns (DLTs) permitted at all driveways, (ii) a raised median restricting all driveway traffic to right-in–right-out and U-turns permitted at signallised intersections, (iii) a peak-hour DLT restriction at all driveways, and (iv) dynamic restriction (i.e. a restriction enforced during the time intervals in which traffic flow rates exceed given thresholds) of driveways to right-in–right-out in a CV environment. On the basis of the simulation analysis, it was found that converting driveway access from fully open to right-in–right-out based on prevailing traffic conditions in a CV environment can improve traffic operations.

Short-term traffic flow forecasting has been regarded as essential for intelligent transportation systems, including both point prediction and interval prediction. Compared with point prediction, interval prediction of traffic flow in the future will be critical for traffic managers to make reasonable decisions. This study applies the fuzzy information granulation method to obtain the dispersion range of the collected traffic flow time series, and classical forecasting approaches of K-nearest neighbours, back-propagation neural network, and support vector regression are applied on the dispersion range and the original series itself, constituting a short-term traffic flow forecasting system with the capability of joint point and interval prediction. Using real-world traffic flow data collected from a field transportation system in America, the proposed forecasting system is shown to generate workable point prediction and associated prediction interval, demonstrating the validity of the proposed forecasting system. In addition, for unravelling the impact of time interval on the forecasting system, different time intervals are investigated, showing that with the increase in time interval, the stability of the forecasting system increases. Discussions are provided for the proposed approach, and future works are expected to enhance the proposed forecasting system.

Rich and complete data play a fundamental role in intelligent traffic management and control applications. A great volume of missing data is found in the intelligent transportation system. In this paper, the authors introduce an ensemble strategy to handle the missing values. The proposed strategy is a general framework that different models, whether linear, neural networks, or other, can be applied. In this strategy, missing values are first computed by the forward and backward models, and their results are combined to recover the incomplete raw data. Then, the models are iterated for several times to enhance the accuracy. Three commonly used imputation models are tested in the proposed strategy using the data from real world. The results indicate that the proposed strategy can significantly improve the accuracy of the imputation with different missing types and during different traffic states. Moreover, the increase of the iteration is capable to improve the performance of the models.

Signal control strategy has a huge effect not only on intersection operations but also on traffic emissions. The basic assumptions are first proposed. Using the second-by-second data of vehicular velocity and acceleration, the mathematical expressions are then presented to calibrate the emission factors during green/red on the basis of vehicle-specific power. Given some necessary constraints, a single-objective optimisation model and a bi-objective optimisation model are formulated with the concern of traffic emissions. Considering three levels of traffic demands and two methods of signal timing, the numerical examples are carried out by utilising the VISSIM and MATLAB software packages. The findings indicate that the emission factors during green are explicitly greater than those during red and they are all stable for each pollutant and for each lane group; a scientific signal control system can simultaneously reduce vehicle delay and traffic emissions for isolated intersections.