It has been previously documented that under the condition of crossing an uncontrolled intersection, the decision-making process of drivers is rather complex and highly interactive: they need to decide about the timing, direction, and strategy to complete the required maneuver to avoid potential conflicts. Given the significant impact of this behavior on intersection safety and capacity, we model the merging behavior of a straight-moving vehicle facing a turning vehicle at an uncontrolled intersection in China. We expect that this model can predict the probability that a straight-moving driver has a preemptive status (i.e., arrive at the merging point before the turning vehicle). The factors which determine the drivers’ decision behavior are analyzed, and then we propose a logistic regression model using a dataset collected from an uncontrolled intersection in Kunming, China. Our model reveals that the speed difference, the distance between the two vehicles, and the distance of the turning vehicle to the merging point are the major determinants of a straight-moving driver’s decision. In addition, the prediction results from our model are compared with real-world observations, and better validate the decision behavior model in China.

The objectives of this paper are to identify the contributing factors to households’ choice of the number of vehicles used for evacuation and to develop predictive models of this choice that explicitly consider the constraint imposed by the number of vehicles owned by the household. This constraint is not accommodated by regular ordered response logit models. Data comes from a poststorm survey for Hurricane Ivan. Two models that are variants of the regular Poisson regression model are developed: a Poisson model with exposure and right-censored Poisson regression. The right-censored Poisson model is preferred due to its inherent capabilities, better fit to the data, and superior predictive power. The model and individual variable analyses indicate that households traveling longer distances or evacuating later are more likely to use fewer vehicles. Households with prior hurricane experience and pet owners are more likely to use a greater number of vehicles. Income and distance from the coast are insignificant in the multivariable models, although the individual effect of distance from the coast has a statistically significant bivariate relationship with vehicle usage choice based on the Pearson correlation measure. A method for using the right-censored Poisson model to produce the desired share of vehicle usage is also discussed for generating individual predictions for hurricane evacuation demand simulation.

The use of Bluetooth detection systems for collecting travel time data is becoming increasingly popular because data can be collected continuously at a relatively low cost. However, representativeness of the data can be problematic, given the relatively low detection efficiency of Bluetooth monitoring systems. This paper investigates the use of multiple sensors with individual antennas to the same Bluetooth data collection device to increase the number of identifications from passing vehicles at each data collection point. Increased detection potentially increases the number of travel time measurements, increasing the potential for the measurements to better represent the travel conditions on the roadway. To test this approach, devices with up to 10 Bluetooth sensors were deployed concurrently on a freeway overpass. The marginal benefit of additional Bluetooth sensors in the array was investigated. The additional detections captured by adding up to six Bluetooth sensors in a single sensor array outnumber the detection losses associated with possible radio interference between the Bluetooth sensors or in device-to-sensor communication. On average, additional Bluetooth device media access control (MAC) addresses were detected by increasing the number of sensors from one to two (61% increase), two to three (27% additional increase), three to four (16% additional increase), four to five (11% additional increase), and five to six (8% additional increase). Also, it was observed that additional travel time measurements were collected by adding the number of sensors from one sensor at both locations to a two-sensor setup at both locations (between 110 and 156%), from a two-sensor setup to a three-sensor setup (between 44 and 60%), from a three-sensor setup to a four-sensor setup (between 30 and 47%), from a four-sensor setup to a five-sensor setup (between 10 and 20%), and from a five-sensor setup to a six-sensor setup (between 8 and 17%). This field study indicates that there are potential benefits to using multisensor Bluetooth arrays in lieu of the current practice of using a single sensor configuration to capture additional travel time data from passing vehicles.

This paper addresses the road toll pricing and capacity investment problem in a congested road network in a multicriteria decision-making framework. A goal programming approach is used in which the following four major goals are considered: (1) cost recovery, (2) service level, (3) environmental, and (4) equity. The multiobjective road toll pricing and capacity investment problem is formulated as a bilevel goal programming model. The upper level of the model aims to minimize the deviations from stated goals for a given priority ranking of the goals, while the lower level is the road users’ route choice equilibrium problem. A simulated annealing-based solution algorithm is developed to solve the proposed model. Numerical results show that the priority structure of the goals can significantly affect the road toll pricing and capacity investment decisions. The simulated annealing-based solution algorithm outperforms the sensitivity analysis-based solution algorithm in terms of solution quality. The proposed methodology provides an avenue for understanding the trade-offs among conflicting objectives and for designing a financially and environmentally sustainable transportation system.

To improve customer satisfaction and reduce operation costs, transit authorities have been striving to monitor transit service quality and identify the key factors to enhance it. The recent advent of passive data collection technologies, e.g., automated fare collection (AFC) and automated vehicle location (AVL), has shifted a data-poor environment to a data-rich environment and offered opportunities for transit agencies to conduct comprehensive transit system performance measures. However, most AFC and AVL systems are not designed for transit performance measures, implying that additional data processing and visualization procedures are needed to improve both data usability and accessibility. This study attempts to develop a data-driven platform for online transit performance monitoring. The primary data sources come from the AFC and AVL systems in Beijing, where a passenger’s boarding stop (origin) and alighting stop (destination) on a flat-rate bus are not recorded. The individual transit rider’s origin and destination can be estimated by utilizing a series of data-mining techniques, which are then incorporated into a regional-map platform for transit performance measures. A multilevel framework is proposed to calculate the network-level speed, route-level travel time reliability, stop-level ridership, and headway variance. These statistics are interactively displayed on a map through a simplified transit GIS data model. This platform not only serves as a data-rich visualization platform to monitor transit network performance for planning and operations, it also intends to take advantage of e-science initiative for data-driven transportation research and applications.

Safety-based volume warrants for right-turn lanes (RTLs) at uncontrolled major-road approaches to intersections and driveways on two-lane roadways were developed based on traffic crash and conflict analyses. Crash analysis was carried out based on statewide historical data of traffic crashes reported on Minnesota’s two-lane trunk highways. Conflict analysis was carried out using field data and twelve individual VISSIM models involving 350 scenarios to simulate the conflicts caused by right turns for a wide range of conditions. Each scenario involved twenty simulation runs, and each run used a different random seed. Model calibrations involved replication of spot-speed and time-headway distributions observed at field, whereas the model validations involved comparisons between simulated and observed conflicts. Crash-conflict relationships were developed in terms of crash-conflict ratios (CCRs) and crash estimation factors (CEFs). The estimated mean CCRs were $1.413\times {10}^{-6}$ and $0.496\times {10}^{-6}$ at approaches without and with RTLs, respectively. The values of CEFs, determined as a ratio of CCR to the probability of rear-end/same-direction-sideswipe crashes caused by right-turning vehicles, were between $0.900\times {10}^{-6}$ at approaches with RTLs and $2.168\times {10}^{-6}$ at approaches without RTLs. The warrant guidelines, in terms of volume thresholds, were established for intersection and driveway approaches separately through benefit-cost analysis. The thresholds for driveway approaches were 10 to 26% lower.