Journal of Transportation Engineering, Part A: Systems

This study analyzed a last-mile fresh food delivery system for individuals in underserved communities. Five alternative delivery networks with various modes were considered: conventional trucks, electric cargo bikes, third-party deliveries by personal car, personalized ride transportation services, and parcel lockers. The total cost was formulated and optimized for each alternative. Then, the optimized results for the alternatives were compared. For our baseline values, numerical results showed that delivery by third party was the most cost-effective alternative in delivering fresh items, while truck deliveries ranked second. Personalized ride service and parcel locker delivery were more expensive than home-delivery services. Although more vehicles and frequent trips were needed for all alternatives as service area increased, home-delivery services had a moderate increase in total cost compared to other types. The personalized ride scenario was less influenced by changes in user value of waiting time. At a low demand density, truck operation may not be economically feasible. The study examined whether truck delivery could perform multiple tasks (i.e., delivering items to customers and fulfillment centers in a single delivery tour). Finally, mode shares for home deliveries were estimated when multiple delivery alternatives coexisted.

This paper describes the development of a breakdown probability model for a long-term work zone in a suburban area. It was found that the Gompertz distribution best describes the breakdown probability at a work zone. In the absence of work zone breakdown probability models, deterministic capacities may still be used for work zone planning because the work zone data sets needed for the construction of breakdown probability models are scarce. The capacity values and the corresponding breakdown probabilities associated with 10 common definitions of deterministic work zone capacity were determined. This information will assist practitioners in choosing work zone capacity definitions that correspond to an acceptable risk of congestion within their jurisdictions. A data mining method was proposed to automate the process for identifying traffic breakdown, and metrics were proposed to assess goodness of fit of various breakdown probability models.

Most studies of road travel time estimation have been based on traffic-flow theory or data-driven methods and generally neglect the influence of urban built environment on road travel time. A global regression model and a geographically weighted regression model were thus established to analyze the spatial heterogeneity of the effects of urban built environment on road travel time. The estimated results of the global regression model indicate that the occupancy rate of taxis, the distance from the nearest intersection, and the speed limit show positive correlations with a road’s travel speed, whiles the number of bus stops and the distance from the nearest school show negative associations with the travel speed of the road. Furthermore, based on the results of the geographically weighted regression model, the spatially varying relationships between urban built environment and road travel time can be established, thus providing important information for decision-makers to reduce road travel time by adjusting the attributes of urban built environment.

As explained in the Highway Capacity Manual (HCM), if the segment length is greater than the maximum weaving length (${\mathrm{L}}_{\mathrm{wmax}}$), it is considered to be separate segments of merging and diverging; otherwise, it is a weaving segment. This article focuses on a situation where the maximum weaving length in merging/diverging segments is equal to the length of the segment or the sum of acceleration and deceleration. In both situations, conditions such as length and traffic volumes are equal. Data analysis in Freeval 2015e shows that merging/diverging segments perform better and follow a regular order in terms of speed and density under equal conditions. Nearly in all cases, the density decreased and the speed increased by increasing volume ratio (VR). However, for the weaving segment, density and speed follow an oscillating order. In addition, this article introduces a variable in ${\mathrm{L}}_{\mathrm{wmax}}$ equation called FR is equal to the ratio of the freeway-to-ramp traffic volume over the ramp-to-freeway traffic volume ($\mathrm{FR}={\mathrm{V}}_{\mathrm{FR}}/{\mathrm{V}}_{\mathrm{RF}}$). A West Hakim Highway weaving segment in Tehran, Iran, was simulated in the capacity conditions with some traffic scenarios, implying maximum weaving length and flow. Analysis of exploratory data showed that $\mathrm{FR}=1.1$ is a threshold that differs in traffic behavior before and after that value, so two regime models were used. Linear regression models for ${\mathrm{L}}_{\mathrm{wmax}}$ were developed in two separate parts of $\mathrm{FR}\le 1.1$ and $\mathrm{FR}>1.1$ for two different situations of ${\mathit{N}}_{\mathit{wl}}=2$ and 3 (${\mathit{N}}_{\mathit{wl}}$ = number of weaving lanes). Results showed that FR should be added to HCM’s ${\mathrm{L}}_{\mathrm{wmax}}$ equation.

Conflicts between the right-turn vehicles and crossing pedestrians at signalized intersections not only endanger pedestrians, but also impede the ability of vehicles to execute their runs. The strategy proposed in this study recognizes the fact that the pedestrian volume in the late stage of the pedestrian green (PG) time is often low. Thus, it terminates the PG in advance to provide the right-turn vehicles with a pedestrian-free interval, which quickens the turning movements and improves the right-turn capacity. A rule-based algorithm is developed to determine the PG termination time dynamically, which considers the priority needs of vehicles and pedestrians. The proposed strategy is validated numerically at an intersection in Shanghai, China. The results demonstrate that the proposed strategy could efficiently improve the right-turn capacity and reduce vehicle delays with limited impact on pedestrians. Sensitivity analyses suggest that the proposed strategy is applicable at signalized intersections with high right-turn vehicle volume and moderate pedestrian volume.

Over the last few decades, there have been two main streams of data used for driving behavior research: trajectory data collected from the field [such as using video recordings and global positioning systems (GPS)] and experimental data from driving simulators (where the behaviors of the drivers are recorded in controlled laboratory conditions). Previous research has shown that the parameters of car-following models developed using simulator data are not directly transferable to the field. In this research, we investigate the differences in detail and compare alternative methods to overcome the problem. Two types of approaches are tested in this regard: (1) econometric approaches for increasing model transferability—Bayesian updating and combined transfer estimation—and (2) joint estimation using both data sources simultaneously. Car-following models based on a stimulus-response framework are developed in this regard, using experimental data collected at the University of Leeds Driving Simulator (UoLDS) and detailed trajectory data collected at California Interstate 80 (I-80), in the US, and the UK Motorway 1 (M1). The estimation results of the initial models show that car-following models using driving-simulator data are closer to the UK (M1) data than the I-80 data but not directly transferable. Performances of the proposed approaches for improving transferability are evaluated using t-tests for individual parameter equivalence and transferability test statistics (TTS). The results indicate that the transferability can be improved after parameter updating, and the combined transfer estimation is found to outperform the other approaches. The findings of this study will enable a more effective usage of the driving simulator data for the estimation of mainstream mathematical models of driving behavior while the techniques used can be applied to other types of econometric models.

The majority of existing evacuation models neglect the fact that leaking toxic gas usually degrades atmospheric visibility under chemical disasters. This paper takes the impact of leaking toxic gas on atmospheric visibility into account in evacuation problems. The reduction of traffic speed and road capacity under conditions of low atmospheric visibility is introduced and has stochastic features. Further, a robust evacuation traffic management model for evacuation scenarios of low visibility and toxic-gas-leakage disasters is developed to obtain optimal evacuation plans. The proposed robust optimization model can guarantee the feasibility of optimal evacuation plans to the evacuation scenarios of low atmospheric visibility under corresponding robustness requirements. Numerical experiments show that neglecting the impact of low atmospheric visibility overestimates the performance of optimal evacuation plans and even assigns evacuation traffic to roads with bad driving conditions under chemical disasters.

In this paper, we investigated the influence of the cross-sectional characteristics of bicycle lanes on motorized and nonmotorized traffic safety. The safety effects of bicycle lanes with different widths were assessed through the estimation of safety performance functions (SPFs) and crash modification factors (CMFs). Four different types of crash prediction models [i.e., generalized linear model (GLM), random effects GLM, generalized nonlinear model (GNM), and random effects GNM] were examined to develop more reliable SPFs. The results indicated that the random effects GNMs provided the most reliable estimates. The goodness of fit was higher for the GNMs than for the GLMs because of the nonlinear relationship between the width of the bicycle lanes and the crash rates. In addition, the random effects models indicated better performance than the GLMs and GNMs. The results indicated that the installation of bicycle lanes is an effective safety measure to reduce four different types of crashes in general. The results of the estimated CMFs using the random effects GNMs indicated that the safety effects of bicycle lanes had nonlinear variations based on different widths. In general, the installation of bicycle lanes with widths of 1.5–1.8 m (5–6 ft) was most effective for reducing motorized crashes, whereas bicycle lanes with widths of 1.8–2.1 m (6–7 ft) were more appropriate for reducing nonmotorized crashes. According to the findings of this study, specific guidance on minimum bicycle lane widths for various roadway characteristics, traffic flows, and roadway types based on empirical evidence regarding traffic safety can be provided.

If connected and autonomous vehicles use highways under current lane management policy, only partial benefit can be reaped from the platooning of advanced vehicles. Agencies may have to update the existing policy in the near future because of the regular presence of connected and autonomous vehicles on the highways. The probable management approaches are likely to affect highway construction, rehabilitation, and fuel consumption cost. This paper studied a range of lane management approaches in terms of truck platooning and estimated construction and rehabilitation costs using existing models. It also developed a probabilistic approach of generating varying degrees of vehicle platooning to estimate fuel consumption in a simulated environment of highway segment. The results showed that, compared with a no-platoon approach, most of the approaches considered reduce construction cost but increase rehabilitation cost for a unit length of a highway. Fuel consumption cost decreases due to truck platooning with all possible approaches. If truck platooning is achieved by adding a separate truck-only lane, the fuel consumption benefits are considerably higher.

Steep downgrades are known as a part of mountain roads with a significant number of severe crashes. Generally, using speed limiters such as speed bumps and speed humps are considered the most effective (considering both the time and cost) ways to reduce the number of victims. The efficiency of speed limiters depends on the driver’s consciousness and the car’s braking system; the latter could be problematic in wet and dark conditions of the road. The main goal of the current study is to propose an approach to correct the longitudinal profile of roads by replacing the consecutive downgrades and upgrades with a continuous downgrade. The vehicle’s speed is reduced by using this method before entering an accident-prone site without the driver’s involvement and breaking. To this end, a well-known software program is used under two speeds (70 and $100\mathrm{km}/\mathrm{h}$) in 68 different routes to analyze speed along the proposed routes. Also, another computer model is used for dynamic simulation. Dynamic analyses and regression models show that the vehicle’s speed in the new proposed route is significantly reduced compared with the continuous downgrade. Moreover, results indicate that the upgrade slope of the proposed route has the most impact on reducing the final speed of the vehicles.

Adaptive cruise control (ACC) vehicles are the first step toward comprehensive vehicle automation. However, the impacts of such vehicles on the underlying traffic flow are not yet clear. Therefore, it is of interest to accurately model vehicle-level dynamics of commercially available ACC vehicles so that they may be used in further modeling efforts to quantify the impact of commercially available ACC vehicles on traffic flow. Importantly, not only model selection but also the calibration approach and error metric used for calibration are critical to accurately model ACC vehicle behavior. In this work, we explore the question of how to calibrate car-following models to describe ACC vehicle dynamics. Specifically, we apply a multiobjective calibration approach to understand the trade-off between calibrating model parameters to minimize speed error versus spacing error. Three different car-following models are calibrated for data from seven vehicles. The results are in line with recent literature and verify that targeting a low spacing error does not compromise the speed accuracy whether the opposite is not true for modeling ACC vehicle dynamics.

The Safe Routes to School (SRTS) program apportions funding for safety interventions aimed at encouraging walking and bicycling as viable transportation modes to school. Initially, the SRTS program (2006–2012) allocated more than $20 million to schools throughout Indiana for infrastructure- and noninfrastructure-related safety interventions. Under the continuation of the SRTS program, many states (including Indiana) do not provide special consideration for SRTS using federally allocated funds. Nevertheless, there are still provisions for SRTS projects to be funded through a more limited federal funding process. Using econometric modeling techniques, this paper examines the initial implementation of SRTS to gauge the effectiveness of infrastructure and noninfrastructure safety interventions. This analysis is beneficial in understanding the program’s most effective measures, and may assist in navigating the more competitive federal funding process by developing guidance on which to focus funding going forward. The impact of SRTS interventions on child (6–17 years) pedestrian and bicyclist crashes near schools was evaluated over time using a panel data structure that included SRTS and control group (no interventions) schools in Indiana. In the period before implementing the SRTS interventions, the schools selected for the program experienced higher crash frequencies than the control group, thus supporting their inclusion in the program. After the program’s implementation, infrastructure interventions were found to be effective in reducing child pedestrian and bicyclist crashes, while noninfrastructure interventions showed a nonsignificant impact. Covariates such as vehicle kilometers traveled (VKT), school enrollment, median age of residents, median income of household head, and average annual precipitation also influenced safety in the region surrounding the studied schools. Based on the results of this study, SRTS programs including infrastructure-related safety interventions appeared to be most promising in improving safety for child pedestrians and bicyclists. Therefore, this study’s findings indicate the expansion of funding to infrastructure-related projects in future phases, as conceived in the initial structure of the SRTS program.

Traffic signal retiming involves periodically updating existing signal coordination plans of signalized corridors. Nevertheless, there are still several limitations with the current signal retiming practice, such as managing a large set of signal coordination data, optimizing coordination parameters, diagnosing timing errors, and assessing signal coordination performance. To fill these gaps, this research introduces a novel signal retiming approach that offers transportation engineers an alternative way to assess and improve the operation of existing coordinated corridors. It allows for displaying the vehicle trajectories in the time-space diagram (TSD), based on which, transportation engineers could diagnose the potential issues with a signal coordination plan and develop an optimized signal coordination plan for the corridor. In addition, a unique signal performance index (SPI), which took into account average travel speed, number of stops, cycle length, and segment length adjustments, was developed to grade the performance of a signal coordination plan. A total of 38 signal coordination plans were applied on 8 signalized corridors in the City of Reno, Nevada, based on the proposed signal retiming approach. Through a before and after study, this research concluded that traffic signal retiming considerably improved the travel speed of the corridors and reduced delay for coordinated movements at signals. An SPI level of 76% of the retimed signal coordination plans was increased, and 16% of the retimed plans witnessed increases in either speed score or stop score. These case studies support that the proposed signal retiming approach is applicable for field practices.

This note documents motorist behaviors when overtaking cyclists. Observers reviewed nearly 2,300 h of video and classified 20,734 vehicle–cyclist interactions on a variety of commonly used bicycle facilities on local streets at nine locations in Minnesota. Descriptive statistics and a multilevel logistic regression model indicate motorist behaviors are more consistent on roadways with facilities that are visible and distinctly marked. When passing cyclists on streets with no facilities, bikeway signs only, or sharrows, drivers are more likely to enter into adjacent lanes than when passing on streets with striped or buffered bike lanes. Regression results indicate no statistically significant differences between motorist behaviors on roadways without any facilities and streets with either sharrows or signs affirming shared use. The results add to evidence that introduction of bike lanes on roadways may reduce behaviors that pose risks.