One of the most critical functions of an intelligent transportation system (ITS) is to provide accurate and real-time prediction of traffic condition. This paper develops a short-term traffic condition prediction model based on the $k$-nearest neighbor algorithm. In the prediction model, the time-varying and continuous characteristic of traffic flow is considered, and the multi-time-step prediction model is proposed based on the single-time-step model. To test the accuracy of the proposed multi-time-step prediction model, GPS data of taxis in Foshan city, China, are used. The results show that the multi-time-step prediction model with spatial-temporal parameters provides a good performance compared with the support vector machine (SVM) model, artificial neural network (ANN) model, real-time-data model, and history-data model. The results also appear to indicate that the proposed $k$-nearest neighbor model is an effective approach in predicting the short-term traffic condition.

This paper presents the results obtained from the estimation of free-flow speed on two-lane rural highways. The data used for the analysis were collected in Northwest Italy using video cameras and a laser speed gun. The model structure adopted separates the estimate of the central tendency of speeds from the typical deviations of individual speeds. Hence, in the model, the same set of variables can be used to determine both the mean value and the standard deviation of the speed distribution; the desired speed percentile is then calculated by considering the associated standard normal random variable ($Z$). Random effects (RE) were included in the model to account for the variability in time and space of the data that contain multiple measurements for the same road/section/direction and to remove any dependency between estimation errors from individual observations.

The research uses a database of urban noise data collected continually from April 2013 to March 2014 at 10 sites in Dublin, Ireland. The first objective of the paper is to investigate if the morning daily noise level peak is related to transport characteristics of households, such as car ownership levels, the mode by which people travel to work, and morning work trip departure time. Data from the 2011 Irish census is used to provide the information on households and this is tested against noise measurement levels. The second objective is to examine the relative importance of the spatial and temporal variables of location, month of the year, weekday, and hour of the day in predicting urban noise levels using a multinomial logistic regression model. The results show that the transport household characteristics examined do not appear to influence noise levels. The outcome from the regression model demonstrates that location is the most important variable followed in order by hour of the day, month of the year, and weekday in predicting urban noise levels.

Tours are increasingly being considered as an appropriate unit of observation of mobility behaviors and are one of the key ideas underpinning contemporary activity-based modeling approaches. Identifying typologies of tours would benefit both modelers and decision makers, striving to set up more tailored actions aimed at promoting environmentally benign travel choices. Different a priori classifications based on activity kinds have been proposed, none of which seems clearly preferable on empirical grounds. This paper takes a complementary approach and defines a data-driven segmentation through a cluster analysis of tours that were derived from the trip records from a United States national survey. The socioeconomic characterization of each cluster is finally carried out to link travelers’ profiles with specific kinds of tours. Four main tour clusters have thus been identified: nonwork tours for compulsory activities done by young individuals, tours done by elder or retired persons, short and secondary tours within the travel day, and tours dominated by the working activity. Their relevance on a modeling and policy viewpoint is discussed.

The heteroscedasticity refers to a collection of random variables with a subpopulation that have different dispersions from others. The variable dispersion could be quantified by measures of statistical dispersion such as standard deviation or coefficient of standard deviation. This study aims to model the effects of rainfall intensity on the heteroscedastic traffic speed dispersion on urban roads. The traffic and rainfall intensity data were collected by a selected video traffic detector and its nearest rainfall station in Hong Kong, respectively. The coefficient of variation of speed (CVS) was employed to measure the vehicular traffic speed dispersion. The analysis shows that the empirical values of CVS typically range from 0.05 to 0.2 at different traffic densities and rainfall intensities, and the exponential function provides a good fit to traffic speed data under both dry and rain conditions. A generalized function of CVS with the effects of rainfall intensity is proposed, calibrated, and validated with different sets of empirical data. The calibration and validation results show that the proposed generalized function of CVS fits well with the empirical data. The empirical findings and the generalized function of CVS proposed in this study may benefit for assessing and modeling the level-of-service performance of urban roads in Pacific Rim cities similar to Hong Kong with relatively high annual rainfall intensity.

Pavement management systems rely on accurate distress measurements to support transportation agency officials in making decisions on budget planning and allocation as well as on the design of maintenance and rehabilitation strategies. Errors in distress data measurements lead to inappropriate project prioritization and increased maintenance and rehabilitation costs. This paper presents an independent evaluation of the accuracy and precision of high-speed field measurements of pavement cracking taken by three different automatic three-dimensional (3D) systems that represent the state of the practice. Cracking data were collected from a field experiment designed to represent typical conditions encountered in Texas highway network and comprised twenty 550-ft pavement sections. Three vendors participated in the experiment; two of them used the same hardware (i.e., INO LCMS sensors) but different proprietary algorithms to detect and quantify surface distresses. The third vendor used PaveVision sensors. The high-speed measurements were compared to manual measurements taken statically by experienced certified raters. Further, each service provider was asked to report their results with different levels of manual intervention in order to capture the change in accuracy due to manual postprocessing of the data. The large number of false positives and low overall accuracy and precision showed by the three automated systems highlights the need for improvements on the available high-speed data collection technologies. The accuracy, along with the amount of false positives, significantly improved after applying manual postprocessing. The precision, on the other hand, was similar. In addition, a weak association between cracking measurement accuracy and surface texture was observed for every automated system.