Journal of Transportation Engineering, Part A: Systems

In this study, the 85th and 98th percentile speed prediction models of car, light, and heavy commercial vehicles for four-lane divided horizontal curves in rural highways were developed using a stepwise multiple linear regression method. In total 10 sites were considered for developing the models and three for validation. Geometric features of horizontal curve and adjacent tangent sections along with vehicle speed at point of curvature and 50 m prior to point of curvature were considered as independent variables. The 85th and 98th percentile speed of car and sport utility vehicle were found to be comparable, and hence combined for further analysis. The 85th percentile speed of commercial vehicles at midcurve were found to be dependent on its speed at preceding section, which was finally predicted based on geometric features considered. For car, it was dependent on curve length and deflection angle. The 98th percentile speed at midcurve of all vehicle types were dependent on its 85th percentile speed. In validation, the root-mean square errors for the developed models were within $10\mathrm{km}/\mathrm{h}$. The study revealed that curve length could be an important factor in the geometric design procedure.

Drivers who make wrong-way entries onto freeways pose a serious risk to the safety of other motorists and themselves. As a new countermeasure to mitigate the wrong-way entry issue, directional rumble strips (DRSs) were designed to generate elevated noises and vibrations to warn against wrong-way drivers and a normal level of stimuli to slow down right-way traffic. Five conceptual designs were developed based on Department of Transportation (DOT) guidelines, existing transverse rumble strips implementations, and input from rumble strip vendors. A national survey and extensive field tests were performed to verify the effectiveness of the proposed configurations. Acoustic and tactile signatures of the DRSs were measured by a specially equipped passenger car under different speed categories. The results indicated that the tested patterns could provide similar sound and vibration levels in the wrong-way direction as the existing transverse rumble strips (61.8–80.0 dBA sound signals and 1.1–1.4 g vibrations). The statistical and comparative analyses identified three DRS configurations that could produce greater audible and tactile signals in the wrong-way direction than the right-way direction, thereby serving the purpose of alerting inattentive wrong-way drivers while offering good visual attentiveness and applicability.

Although pedestrian countdown signals (PCSs) are meant for pedestrians, they give cues to drivers on the remaining amount of green as the timer counts down. This study focuses on the evaluation of safety effectiveness of PCSs to drivers in the cities of Jacksonville and Gainesville, Florida, using the before-after study with the empirical Bayes method. This analysis explored 110 intersections with PCSs and their respective 93 comparison sites. The findings indicate that PCSs significantly improve driver safety by 8.8% reduction in total crashes, 8.0% in rear-end and 7.1% in property-damage-only crashes, where both of these results were significant at the 95% confidence level. Results for angle crashes as well as fatal and injury crashes were not significant at the 95% confidence level. Also discussed in this study are the crash modification functions developed to show the relationship between the estimated crash modification factors and total entering traffic volume at the intersection. In summary, the results suggest the usefulness of PCSs for drivers.

This paper aims to evaluate travel time variability as well as reliability indexes using global positioning systems (GPS)–based trajectory data of bus trips collected along a selected bus route of the city of Chennai in the southern part of India. Travel time reliability indexes, such as planning time index (PTI), buffer time index (BTI), and buffer time (BT), along with other statistical measures over different time periods are estimated. Generalized extreme value (GEV) distribution is found to be the best-fitted distribution for explaining bus travel time variability reasonably well, using the Kolmogorov-Smirnov (KS) test. Buffer time and 95th percentile travel time are the reliability measures with the most potential, the variation of which reasonably matches the variation in $k$-value (shape parameter of GEV distribution) over time. The findings from the statistical distribution analysis indicate that travel times during peak hours can be better described using normal distributions. The generic model is developed for predicting volumes based on bus journey speeds. Further, the developed model is validated with the help of travel time data of the same route during a different time period. The study also attempts to demonstrate a methodology for establishing level-of-service (LoS) criteria using reliability indicators. The classification of reliability indicators, considering segment-level travel time data, coefficient of variation (COV) of travel time, and volume-to-capacity ratio ($V/C$), is finally presented using the cluster technique. Finally, the study concludes that the most effective performance indicators for examining travel time variability on a given bus route are 95th percentile travel time and BT.

This study proposed a quantitative approach to evaluate the effects of dwelling buses on the traffic operations of nonmotor vehicles at stops in China. The primary objectives of this study were to compare the changes in nonmotor vehicle speeds with/without dwelling buses by applying statistical methods; and to develop a generalized event count (GEC) model for traffic conflict prediction and analysis. In addition, a heat map was used in this study to better understand the changes in location of conflicts at bus stops. Data were collected by an unmanned aerial vehicle (UAV) with a visual range of 60 m at the near side, far side, and midblock stops in China. The results showed that the average effect, which was estimated by averaging the differences over all three types of bus stops, was $(2.5+4.1+1.8)/3=2.8\mathrm{km}/\mathrm{h}$. Furthermore, to evaluate the performance of the proposed GEC model, another model based on the queuing theory method was used for comparison. According to the results, the GEC model had more accurate and reliable fitted values [with 12.80% of mean absolute percentage error (MAPE) and 0.8442 of R-squared] than the compared method (with 18.55% of MAPE and 0.7397 of R-squared). Consequently, with the proposed method, it was feasible to evaluate the effects of dwelling buses on the traffic operations of nonmotor vehicles at bus stops. Finally, to improve the traffic operations at stops, transportation agencies could consider implementing countermeasures to control the conflicts between dwelling buses and nonmotor vehicles, such as reducing bus delay time at stops.

A new approach for calibrating Robertson’s platoon dispersion model (RPDM) is proposed for addressing the issue of overestimation of time span of downstream vehicle arrival as predicted by the model Instead of comparing the observed and predicted vehicle arrivals during different modeling time intervals, the new approach for calibrating RPDM concentrates on the time span of downstream vehicle arrivals to model the vehicle platoon dispersion. The new approach is demonstrated using vehicle platoon data collected from a signalized urban arterial. Vehicle arrival profiles as obtained using the new approach for calibrating RPDM and the traditional approach are further compared. The error related to overestimation of time span of downstream vehicle arrival is found to be lower for the new approach for calibrating RPDM than the traditional approach. Microsimulation models were used to evaluate the implications of the new approach to calibration on signal design. Three scenarios (SC-I, SC-II, and SC-III) were hypothesized with variations in link length between a pair of coordinated traffic signals. All three scenarios resulted in the reduction of vehicle delay along the coordinated links while adopting the new approach for calibrating RPDM to develop the signal timing plan using TRANSYT-7F. Although these are case-specific findings, they encourage further testing of the new approach to calibration on the design of coordinated traffic signal systems for several corridor and traffic conditions.

Elevation information and its derivation, such as grade, are very important in analyses of traffic operation, safety performance, and energy consumption on highways. Google Earth (GE) is considered a reliable source of elevation information of ground surface and highway elevation. Data were extracted from GE. However, the authors found that raw GE elevation data on highways contains various anomalies and noises. The primary objective of this study was to evaluate the use of the ensemble empirical mode decomposition (EEMD) method for anomaly detection and cleaning of highway elevation data extracted from GE. Three interstate highways’ segments were studied, and typical anomalies that existed in raw GE elevation data were identified. The EEMD method was then applied to decompose elevation data into different compositions with different details of original data, which were determined into those containing true information or white noise. The modeling results showed that the EEMD method was effective in excluding noises and obtaining accurate elevation data. Findings of this study can help transport authorities to create an accurate elevation data set for all highways or other road classes.