Soheil Sohrabi   Transportation Researcher

Previous studies showed that optimizing an individual intersection in the network does not result in minimum network delay. This paper aims to provide a timing optimization algorithm for traffic signals using an internal timing policy based on balancing the queue time ratio of vehicles in network links. In the proposed algorithm, the difference between the real queue time ratio and the optimum one for each link of the intersection was minimized. To evaluate the efficiency of the proposed algorithm on traffic performance, the proposed algorithm was applied in a hypothetical network. By comparing the simulation software outputs, before and after implementing the algorithm, it was concluded that the traffic flow after optimizing signal timing increases by 9%, and the total delay in the network decreases by 8%.

This study explores a practical value for highway capacity that is able to represent stochastic capacity and optimize the performance of designed facilities. Optimal performance of roadways could be defined in terms of cost, safety, reliability, and vehicle throughput. We optimize the performance of freeway facilities using the traffic efficiency concept and maximizing vehicle throughput. Traffic efficiency, which is a combination of flow rate breakdown probability and capacity drop, is determined for one of the major freeways in Tehran, Iran. We depict flow-efficiency diagrams for freeway sections to extract the optimum capacity. Comparing the proposed optimum capacity and conventional capacity, we suggest the design value of capacity as 90% of conventional capacity values. The proposed optimum capacity in this study ensures the maximum vehicle throughput in designed facilities.

Public Bicycle Sharing Systems (BSSs) are becoming increasingly popular in recent times. Both the BSS operators and the customers can benefit from the large digital data portals that continuously record the state of the BSS. In this context, two approaches have been proposed for real-time BSS demand prediction:

​

First, we developed generalized extreme value (GEV) count models that can predict hourly bike arrivals and departures at each station while accounting for time-of-day, weather, built environment, infrastructure, temporal, and spatial dependency factors. The proposed models were used to analyze the demand patterns in the Capital Bikeshare system and were found to predict the demand at both aggregates and disaggregate levels with reasonable accuracy. Specifically, the total demand in the entire system was predicted within 5% margin of error whereas 75% of the station-level arrival and departure predictions in the next one hour were within a margin of one from the observed counts. The proposed modeling system is useful (a) to BSS operators to anticipate the future demand and optimize their rebalancing plans, and (b) to BSS customers to better plan their travel based on expected bike and dock availability at the origin and destination ends of their BSS trips.

​

Second, the real-time demand of the system predicted based on the observed demand using machine learning techniques. The proposed model, (1) updates in real-time, (2) captures the effect of weather, infrastructure and socio-demographic variables, and (3) improves the prediction power of the previous models. The model implemented on Washington DC BSS historical data to predict the absolute difference of arriving and departing bikes in station level. Results show that a 95% likelihood of predicting the demand in the next hour with 1 marginal error. Also, the model is able to predict the demand in the next 15-minutes with 1 marginal error with 90% confidence. The Mean Absolute Error of the model is equal to 0.33 (bike).

Enforcing red-light runners is known as an engineering solution to enhance intersection safety. However, the efficiency of automated red-light camera (RLC) programs is always questioned mainly because of the inaccuracy in post-implementation reviews and difficulties with the financial viability of the program. An engineering analysis can address the concerns and improve the RLC programs. In this paper, we propose a methodology to capture the effect of RLCs on intersection safety by including the spatial dependency between intersections, unobserved heterogeneity, and the spillover effect of enforcing cameras. In this context, a Bayesian hierarchical model is implemented to spatially predict crash frequency at intersections. The safety impact of risk factors, such as land-use and intersection function, geometry, and control characteristics, are examined. The proposed model was developed using 150 intersections located in the City of Chicago. The results show that the probability of crashes decreases at intersections equipped with enforcing cameras by 6%. Also, the spillover effect of cameras is confirmed in this study by capturing the safety impact of cameras on other adjacent intersections. It is shown that crash risk is reduced by 2% for an intersection located within a 1 km network distance from the RLC. Defining an optimization problem and employing the findings of this study, we could find the optimal RLC locations across the city that can reduce injury crashes by 13%.

In addition to the basic role of transportation systems to satisfy individuals’ mobility, transportation imposes both positive and negative health impacts in cities. Transportation air pollution exposures are one of the significant causes of deaths where traffic emissions alone are responsible for one-fifth of the Ozone and PM2.5 attributed mortality in the world. According to the World Health Organization (WHO), road crashes are among the top 10 causes of death with 1.4 million road traffic deaths reported globally in 2016. Also, traffic noise is known as a source of environmental noise with a proven adverse effect on public health. On the other hand, transportation systems can also benefit health and encourage users to undertake routine physical activities such as walking and cycling.

Recognizing the impact of transportation on public health and quantifying its burden can directly contribute to (1) transportation planning and infrastructure design, (2) urban planning and design, (3) automotive industry, (4) and health sectors. Considering the health outcomes by transportation planners result in imposing efficient transportation policies such as road-pricing, investing in public transit, encouraging cycling and walking and parking restrictions. Also, safe and environment-friendly infrastructure designs are expected after bringing the impact of the transportation infrastructures (e.g., highways and intersections) into account by transportation engineers. Quantifying the transportation health burden enables urban planners to maintain the land-use balance in the city and develop sustainable transportation infrastructures. The quantified health advantages of electric vehicles and connected automated vehicles encourage the automotive industry to invest more in advancing in safe with zero-emission vehicle designs. The health sectors can detect the high-risk spots in the city and impose policies to avoid further damages and promote well-being. In addition, encountering the transportation impact on public health can indirectly contribute to economic studies and government decision-making such as gas pricing.

The potential detrimental and beneficial impact of transportation on public health has been discussed extensively in the literature and also quantified for an individual risk factor (e.g., noise, air pollution, crashes, and physical activity); however a few studies addressed the integrated disease burden from transportation-related risk factors. Moreover, several quantification tools for Health Impact Assessment (HIA) of transportation policies, plans, or design alternatives exist, namely Health Economic Assessment Tool (HEAT), Integrated Transport and Health Impact Model (ITHIM) [12], and Transportation and Health Tool (THT).

Supporting policies are required to govern the unintended consequences of Autonomous Vehicles (AVs) implementation and maximize their benefits. The first step towards formulating policies is identifying the potential impacts of AVs. While the impacts of AVs on the economy, environment, and society are well explored, the discussion around their beneficial and adverse impacts on public health is still in its infancy. This study provides a review of the literature on AVs and public health and develops a framework to clarify the potential impacts. The proposed model, first, summarizes the potential changes in transportation after AVs implementation into seven points of impacts: transportation infrastructure, land-use, and the built environment, traffic flow, transportation mode choice, transportation equity, jobs related to transportation, and traffic safety. Second, the transportation-related risk factors that affect public health are outlined. Third, we formulate the pathways between AVs and public health using the knowledge gained from two previous steps. The review of the literature shows that the discussion around AVs impacts on public health is gaining increasing research attention in recent years but still needs more attention. Using the proposed model, we found that AVs can be associated with public health through 32 pathways, where they can adversely impact health through 24 of those. The health impacts of AVs are contingent upon supporting policies. Equipping AVs with electric motors, regulating urban area development, implementing traffic demand management, controlling AVs ownership, and imposing ride-sharing policies are the strategies that can reinforce the positive impacts of AVs on public health.

Automotive navigation systems seek the shortest route between a given set of origin and destination points. However, although the suggested routes may help users minimize their travel times, there are certain situations in which the shortest route is not necessarily the safest one. Navigating through local roads that have higher risks of crashes―namely, those with poor geometric designs, drainage problems, lack of illumination, higher risks of wildlife-vehicle collisions, and more interruptions in traffic flow―compared to using higher classification highways is an example of the unintended consequences of routing to ensure minimum travel time. This study examined the problem by comparing the safest and shortest routes between five metropolitan areas in Texas, including more than 29,000 road segments. The study also designed a system architecture for finding the safest route and highlighted barriers to implementing such a system. The results of comparing the safest route and the shortest route between pairs of origins and destinations showed that the shortest route is not necessarily the safest, where an 8% decrease in travel time was associated with a 23% higher risk of being involved in a crash. In addition, the safest route varies according to different weather conditions. The requirements for deploying safety in route-finding systems were identified as (1) availability of real-time traffic flow and incident data for dynamic route-finding systems, (2) more accurate crash prediction models, and (3) a methodology for dealing with the trade-offs between travel time and safety to find the optimal route.

Automotive navigation systems seek the shortest route between a given set of origin and destination points. However, although the suggested routes may help users minimize their travel times, there are certain situations in which the shortest route is not necessarily the safest one. Navigating through local roads that have higher risks of crashes―namely, those with poor geometric designs, drainage problems, lack of illumination, higher risks of wildlife-vehicle collisions, and more interruptions in traffic flow―compared to using higher classification highways is an example of the unintended consequences of routing to ensure minimum travel time. This study examined the problem by comparing the safest and shortest routes between five metropolitan areas in Texas, including more than 29,000 road segments. The study also designed a system architecture for finding the safest route and highlighted barriers to implementing such a system. The results of comparing the safest route and the shortest route between pairs of origins and destinations showed that the shortest route is not necessarily the safest, where an 8% decrease in travel time was associated with a 23% higher risk of being involved in a crash. In addition, the safest route varies according to different weather conditions. The requirements for deploying safety in route-finding systems were identified as (1) availability of real-time traffic flow and incident data for dynamic route-finding systems, (2) more accurate crash prediction models, and (3) a methodology for dealing with the trade-offs between travel time and safety to find the optimal route.

Vehicle automation safety must be evaluated not only for market success but also for more informed decision-making about Automated Vehicles' (AVs) deployment and supporting policies and regulations to govern AVs’ unintended consequences. This study is designed to (1) identify the AV safety quantification studies, evaluate the quantification approaches used in the literature, and uncover the gaps and challenges in AV safety evaluation, and (2) propose new methodologies for AV safety evaluations at the vehicle and society-levels while addressing the identified gaps in the literature.

We employed a scoping review methodology to identify the approaches used in the literature to quantify AV safety. After screening and reviewing the literature, six approaches were identified: target crash population, traffic simulation, driving simulator, road test data analysis, system failure risk assessment, and safety effectiveness estimation. We ran two evaluations on the identified approaches. First, we investigated each approach in terms of its input (required data, assumptions, etc.), output (safety evaluation metrics), and application (to estimate AVs' safety implications at the vehicle, transportation system, and society levels). Second, we qualitatively compared them in terms of three criteria: availability of input data, suitability for evaluating different automation levels, and reliability of estimations. This review identifies four challenges in AV safety evaluation: (a) shortcomings in AV safety evaluation approaches, (b) uncertainties in AV implementations and their impacts on AV safety, (c) potential riskier behavior of AV passengers as well as other road users, and (d) emerging safety issues related to AV implementations. This review is expected to help researchers and rulemakers to choose the most appropriate quantification method based on their goals and study limitations. Future research is required to address the identified challenges in AV safety evaluation.

​

We also proposed a 5-step framework for investigating AV safety at the society level. An empirical analysis was conducted using 2017 crash data from the Dallas-Fort Worth area, the fourth largest metropolitan area in the United States. The results showed that AVs could potentially prevent up to 50%, 46%, 23%, 6%, and 5% crashes for automation levels 5 to 1, respectively. AVs were shown to be more effective in preventing non-injury crashes. Among advanced driver assistance systems (ADASs), pedestrian detection, electronic stability control, and lane departure warning have more significant potential in reducing fatal crashes. We found a U-shaped relationship between preventable fatalities and AVs and household median income and more significant safety impacts on ethnically diverse communities. 

​

In addition, a novel methodology is proposed on the basis of survival analysis methodology which investigates AV safety in comparison with conventional vehicles. AV crash data from the California Department of Motor Vehicles and SHRP2 Natutalisitic Driving Survey are employed in this study. Our results showed that existing Level 3 AVs are safer than conventional vehicles, however, our results carry uncertainties given the limitations in the availability of AV crashes detailed information.

​

​