Safety is perhaps the most important factor in planning, design and operations for all modes in developing countries. Traffic accidents, in particular, constitute the leading cause of death and injury for young people in developing countries, and many of them involve pedestrians and bicyclists. This session explores the many dimensions of this issue.
Adoption of Car Restriction Policies Across 287 Chinese Cities
Shenhao Wang, Massachusetts Institute of Technology (MIT)Show Abstract View Presentation
Jinhua Zhao, Massachusetts Institute of Technology (MIT)
With rapid auto growth, many Chinese municipalities implemented car restriction policies. Research identified several groups of underlying factors that possibly prompted the municipalities to adopt the policies, but no study has yet systematically evaluated these factors. We collected a panel dataset for all 287 Chinese municipalities from 2001 to 2014, extracted information of car restriction policies from 116 legal documents and other online resources. Duration models were used to assess the statistical significance and the predictive accuracy of 14 indicators measuring economic power, population, air pollution, automobiles, urban density, and local transportation conditions in predicting the adoption of large-scale car restriction policies, supplemented by the cross-examination with text analysis of policy documents. We found that the adoption of large-scale car restriction policies primarily responded to air pollution and secondarily to motor per capita and congestion. Policy adoption responded to local subway line constructions, but not other transportation or land use conditions. Local economic power and population size cannot effectively explain policy adoption. Idiosyncratic effects at provincial or city levels were important, although the underlying mechanism was unclear. Broadly, our findings suggest that 1) Chinese municipalities were partially rational in making policies to address local problems; 2) legal documents were reliable to illustrate the real motivations of policies.
Safety Assessment of Two-Lane Highway Using a Combined Proactive and Reactive Approach: A Case Study from Indian National Highways
Sudipa Chatterjee, Indian Institute of Technology, KharagpurShow Abstract View Presentation
Sudeshna Mitra, The World Bank
In India, approximately 30% of road accident fatalities occur on two-lane rural roads. Thus, research focused on identifying risk factors on this road type is of immense interest to most of the road agencies. Although these highways are more hazardous and account to more severe crashes there is a lack in scientific safety assessment on such highways compared to multilane highways. Of late proactive approach such as Road Safety Audit has been widely adopted by the government of India to reduce crash frequency and severity on highways.However, an effective road safety management program should exercise an optimal balance between reactive and proactive strategies to identify potential hazard and to treat the already existing hazardous sites. Through a case study on two 2-lane highways, several risk factors were identified using principles of Road Safety Audit, and were mapped with the available crash data analysis to develop a Risk Matrix. This Risk Matrix was found to be helpful in the selection of countermeasure design in a more scientific way targeting the frequent crash types and severities expected to result at the high crash sites. Finally, it was observed that integrating the findings from reactive analysis with proactive safety management is more beneficial, since they are methodically proven with historical crash records and provide the knowledge of plausible safety hazards at sites where similar features exist. The proposed methodology could be adopted by the road agencies in India and other developing countries for effective proactive safety planning.
Crash Risk Evaluation and Crash Severity Pattern Analysis for Different Types of Urban Junctions: Fault Tree Analysis and Association Rules Approaches
Peijie Wu, Harbin Institute of TechnologyShow Abstract
Xianghai Meng, Harbin Institute of Technology
Li Song, University of North Carolina at Charlotte
Wenze Zuo, Harbin Institute of Technology
Urban junctions usually present a significant safety concern, and a majority of total crashes in urban areas happened in or near the junctions. Efforts have been made in exploring the contributory factors of crash severity at junctions, however, the crash risk levels and crash severity patterns of different junction types were rarely investigated in previous literatures. In order to fill this gap, the safety performance of six junction types and the contributory factors of crash severity were analyzed in this study, which is helpful for city transportation authorities to conduct effective countermeasures. Fault tree analysis (FTA) was applied for the risk evaluation of urban junctions and association rules (AR) algorithm was employed for the crash severity pattern analysis based on the data of STATS19 database from 2012 to 2016. Overall, four types of urban junctions with high crash risk level and over 4,000 association rules contributed to crash severity were identified in the present paper. Results show that 1) roundabouts and mini-roundabouts have the lowest fatality and casualty while T or staggered junction and crossroads have highest crash risk level; 2) FTA may produce inaccurate outcomes because of incorrect logic gates, but RA can generate real potential relationships between crash severity and risk factors; 3) the crash severity pattern is quite complex and the interdependence between risk factors of each junction type is different; 4) risk factors such as male, no physical crossing facilities within 50 meters, and give way or uncontrolled are common in high-risk junctions at night.
Motorcycle Usage Time and Tour Type: Accelerated Failure Time Approach
Amir Bahador Parsa, University of Illinois, ChicagoShow Abstract View Presentation
Arsham Bakhtiari, No Organization
Meeghat Habibian, Amirkabir University of Technology
Abolfazl (Kouros) Mohammadian, University of Illinois, Chicago
By the escalation of dilemmas caused by motorcycles such as air and noise pollution, congestion, and crash injuries, which result in adverse impacts on physical and mental health of citizens, most of the studies in transportation scope have focused on the ownership of motorcycles. However, it seems that these studies have ignored this fact that some people may own more motorcycles while their usage are negligible or vice versa. Therefore, the vitality of conducting a research aiming at motorcycle usage time seems crucial. The aim of the present paper is to identify effective factors on motorcycle usage time of motorcyclists in work-tours. To do so, we conducted a survey of 503 motorcyclists whose workplaces were located in the Central Business District (CBD) of Tehran, as the most polluted and congested area of the city through a face-to-face interview. Based on literature, people's trips are getting increasingly complex, so a tour-based approach is considered in this study. For identifying and comparison of variables reflecting motorcycle usage time in simple and complex work-tours, two Accelerated Failure Time (AFT) models are developed. The results indicate that home location and tour start time are influential in usage time in simple tours while the trip number, motorcycle age and driving experience are effective in complex tours. Moreover, it has been shown that trip distance and monthly fuel cost are the two important factors in motorcycle usage time in both simple and complex tours. However, the effect of trip distance is greater in complex tours.
Gap Acceptance Based Safety Evaluation of Urban Mid-Block Crossings Under Mixed Traffic Environment
Avinash Chaudhari, Research ScholarShow Abstract View Presentation
Ninad Gore, Sardar Vallabhbhai National Institute of Technology
Shriniwas Arkatkar, Sardar Vallabhbhai National Institute of Technology, Surat
Gaurang Joshi, Sardar Vallabhbhai National Institute of Technology
Srinivas Pulugurtha, University of North Carolina, Charlotte
This study aims to evaluate pedestrian safety at urban midblock crossings using gap- acceptance behaviour under mixed-traffic conditions prevailing in a developing country, like India. To achieve the objective, pedestrian-vehicle conflict was analyzed using gap acceptance behaviour at nine urban midblock crossings with varying roadway geomtry, landuse, and traffic characteristics. To evaluate pedestrian safety using gap acceptance behaviour, critical gap was devised as a potential measure. Primarily, accepted gaps for the subject study sections were checked for its distribution and based on goodness-of-fit indices, Generalized Extreme Value (GEV) was found to be the best-fitted distribution among other potential statistical distributions for explaining variation in accepted gaps and, hence, pedestrian crossing behaviour. Thereafter, GEV distribution was extended to obtain probability of crash in accordance to the critical gap values for subject study sections. Later, Post Encroachment Time (PET), a proximal safety concept, is introduced, to validate the crash probability obtained through gap acceptance behaviour. The probability of crash obtained through gap acceptance behaviour and PET were found to be statistically similar at 5% level of significance. Therefore, it can be strongly projected that acceptable gaps can also be considered as a potential safety surrogate measure to evaluate safety at midblock sections. Further, gap acceptance behaviour was extended to delineate gap-acceptance based safety thresholds using k-mean clustering technique, which can facilitate planners and engineers to identify requisite pedestrian facility based on pedestrian crossing behaviour for different roadway, land-use, and traffic characteristics.
Behavioral Evaluation on Different Types of Pedestrian Crossing Facilities: A Case Study in Fortaleza, Brazil
Caio Torres, Universidade Federal do CearaShow Abstract View Presentation
Lucas Tito Pereira Sobreira, Universidade Federal do Ceará
Manoel Castro-Neto, Universidade Federal do Ceará
Flávio José Cunto, Universidade Federal do Ceará
Andrés I. Vecino-Ortiz, Johns Hopkins University
Katharine A. Allen, Johns Hopkins University
Adnan A. Hyder, Johns Hopkins University
Mixed traffic intersections lead pedestrians to share the road with motor vehicles, forcing them to negotiate crossings that often become dangerous. It has become challenging for pedestrians to deal with the ever complex, sometimes hostile, traffic conditions, especially in urban areas that provide no proper pedestrian facilities. This paper provides an exploratory and confirmatory analysis using categorical choice models focused on pedestrian crossing behavior on seven types of crossing control and facilities in the city of Fortaleza - Brazil. The types of control (with or without camera enforcement) and facilities (marked or raised), the characteristics of the pedestrian (age, gender) and its behavior during crossing (speed, distraction, type of crossing and number of people) were evaluated considering crosswalks at mid-block and at signalized intersections. Results showed that pedestrians tend to suffer more delay and to experience riskier situations in midblock marked crosswalks. The existence of exclusive pedestrian interval and camera enforcement decreased the chances of reckless crossings. The logistic regression model for mid-block crosswalks indicates that males tend to behave riskier than females and their likelihood of distraction is considerably higher when crossing in groups. Sites with exclusive pedestrian interval are associated with higher levels of distraction and crossings made in the marked site, compared to raised crossings, were riskier and more aggressive. It was found that the crossings at intersections made during the pedestrian red at sites that provide exclusive pedestrian intervals were also linked to riskier crossings.
Examining the Relationship Between Road Safety Outcomes and the Built Environment in Bogota, Colombia
Erik Vergel-Tovar, Universidad del RosarioShow Abstract View Presentation
Jose Segundo Lopez Valderrama, World Resources Institute
Natalia Lleras, WRI Ross Center
Dario Hidalgo, World Resources Institute (WRI)
Maryfely Rincón, Universidad del Rosario
Sebastian Rincón, Universidad del Rosario
The study of the relationship between the built environment and road safety suggest there are associations between density and urban design features with road safety. This paper conducts a quantitative data analysis with generalized ordinal logit models and log-linear regressions to estimate the influence of the built environment on road safety in Bogota and focuses on road crashes outcomes by estimating the influence of built environment attributes on fatalities and injured victims. The analysis was conducted with georeferenced road crashes data from 2012 to 2016 provided by the Secretariat of Mobility of Bogota. The quantitative data analysis was focused on arterial roads, Bus Rapid Transit System corridors and stations in the city. This analysis was complemented with interviews on site. The results suggest that the presence of pedestrian bridges is positively associated with the number of road crashes. Special attention to road safety in the design of BRT terminals and accesses to stations is needed as the infrastructure might increase speeds and attract pedestrian demand.
Analysis of the Impact of Fog-Related Reduced Visibility on Operating Speed of Multi-Lane Expressways in Shanghai, China
Yanting Liu, Tongji UniversityShow Abstract View Presentation
Yajie Zou, Tongji University
Yanli Wang, Tongji University
Bing Wu, Tongji University
Adverse weather can significantly affect the operation of freeway traffic flow. A lot of studies on the impacts of rainfall, snowfall and other weather conditions on expressway have been conducted. However, few studies focus on the impacts of fog-related reduced visibility on operating speed of multilane expressways in China. This paper analyzed the historical real-time traffic data and meteorological data of 17 months on two typical multilane expressway sections in Shanghai. Multiple factors including type of lanes, type of vehicles, day of the week, and time of day, are considered. The visibility is classified into three levels in all analyses. First, K-S test was used to compare speed distributions under different visibility conditions with factors of type of lanes and another factor in the above multiple factors simultaneously. There are some important findings, e.g. impact of reduced visibility on speed is higher for cars than trucks; it is higher during weekends than weekdays; it is also higher during night than peak hours and daylight. Second, the multiple linear regression model between speed and factors of visibility and factors including lanes, type of vehicles, day of week, time of day are established. The results show all factors are significant. The speed reduces 4.26km/h and 0.56km/h in site 1 and 2, respectively, under low visibility level (<300m). The speed reduces 8.25km/h and 10.07km/h in site 1 and 2, respectively, under moderate visibility level (300-2000m). The developed speed equations will aid traffic management agencies to better design the variable speed limit strategies.
Evaluating Pedestrian Vehicle Interaction Dynamics: A Prudent Approach Using Semi-Automated Safety Diagnosis
Ankit Kathuria, Indian Institute of Technology, RoorkeeShow Abstract
P. Vedagiri, Transportation Systems Engineering Group Department of Civil Engineering Indian Institute of Technology Bombay Powai, Mumbai - 400 076, India. Phone: +91-22-2576 7307 (Off) , 2576 8307 (Res)
Increasing vehicular flow in urban areas has made pedestrians the most vulnerable road users. It is a well-known fact in most of the developing countries that the accident data is not comprehensive. For e.g. in India, the source of accident data is First information report which is not an accident report but a police record of the accidents. Using this data might lead to unsatisfactory analysis and results for research studies. To avoid this, various researchers have demonstrated using surrogate safety measures to evaluate safety of the different types of road facilities. The present research uses surrogate measures by considering an advanced pattern based approach to categorize pedestrian-vehicle interactions based on the road user behaviour. A concept of two interaction pattern has been proposed, Pattern 1 (hard interaction) deals with the responsive behaviour of road user and pattern 2 (Soft/ No Interaction) deals with normal road user behaviour. The behaviour based patterns were categorized based on the Speed, Time to Collision and Gap Time profiles of the pedestrian and vehicle interacting on an un-signalized intersection. In this paper, Import Vector Machine approach has been used to classify the severity levels based on 1021 events extracted from two Un-Signalized intersections in Mumbai, India. Along with the aforementioned categorization an analysis of variable importance was carried out to understand the contribution of different conflict indicators in deciding the severity levels. The severity levels will help to test and evaluate various infrastructure and control improvements for making urban intersections safe for road users.
A Study of Road Traffic Injuries Using Data from Trauma Care Facilities: Additional Perspectives from India
Parag PareekhShow Abstract View Presentation
Sudeshna Mitra, The World Bank
Dipanjan Mukherjee, Indian Institute of Technology Kharagpur
Shirin Wadhwaniya, Johns Hopkins University
Shivam Gupta, Johns Hopkins University
Adnan A. Hyder, Johns Hopkins University
Road traffic injuries (RTIs) are globally recognized as a public health issue with the problem being particularly acute in Low and Middle-Income Countries (LMICs) such as India. This increase in crash frequency and injury severity is linked to construction of high speed corridors amidst poor access control and the concomitant phenomena of rapid urbanization and motorization. While behavior of road users, lack of enforcement, and poor road design have all contributed to the poor road safety scenario, quality of post-crash response also significantly impacts injury severity and deaths. Hence, analysis of RTI data from hospital Emergency Room can yield important insights about critical determinants of post-crash response. This paper presents key findings from a hospital surveillance study with RTI data collected at the Emergency Room (ER) of hospital at the time of admission. Further, the study also compares and contrasts information about injury data collected by police at crash scene. Key findings from the analyses of the emergency room data indicate significantly high representation of pedestrians and motorcyclists amongst all crash victims. Significant associations are revealed between hospital transfer time, mode of transfer, and injury severity. Adverse impacts of not using safety equipment while travelling are also reflected in the analyses. Keywords: Road Traffic Injuries (RTI), Public health, Low and middle-income countries (LMICs), Injury Severity, Post-crash response, Ordered Probit Model