Kevin Majka

Analysis of stationary and dynamic factors affecting highway accident occurrence: A dynamic correlated grouped random parameters binary logit approach

Traditional accident analysis typically explores non-time-varying (stationary) factors that affect accident occurrence on roadway segments. However, the impact of time-varying (dynamic) factors is not thoroughly investigated. This paper seeks to simultaneously identify pre-crash stationary and dynamic factors of accident occurrence, while accounting for unobserved heterogeneity. Using highly disaggregate information for the potential dynamic factors, and aggregate data for the traditional stationary elements, a dynamic binary random parameters (mixed) logit framework is employed. With this approach, the dynamic nature of weather-related, and driving- and pavement-condition information is jointly investigated with traditional roadway geometric and traffic characteristics. To additionally account for the combined effect of the dynamic and stationary factors on the accident occurrence, the developed random parameters logit framework allows for possible correlations among the random parameters. The analysis is based on crash and non-crash observations between 2011 and 2013, drawn from urban and rural highway segments in the state of Washington. The findings show that the proposed methodological framework can account for both stationary and dynamic factors affecting accident occurrence probabilities, for panel effects, for unobserved heterogeneity through the use of random parameters, and for possible correlation among the latter. The comparative evaluation among the correlated grouped random parameters, the uncorrelated random parameters logit models, and their fixed parameters logit counterpart, demonstrate the potential of the random parameters modeling, in general, and the benefits of the correlated grouped random parameters approach, specifically, in terms of statistical fit and explanatory power.

Preliminary Investigation of the Effectiveness of High-Visibility Crosswalks on Pedestrian Safety Using Crash Surrogates

This paper, with the use of data from the SHRP 2 naturalistic driving study, provides a preliminary evaluation of the effectiveness of high-visibility crosswalks (HVCs) in improving pedestrian safety at un-controlled locations. This evaluation was accomplished by analyzing the driving behavior of SHRP 2 participants at three uncontrolled locations at the Erie County, New York, test site. In this context, crash surrogates (i.e., speed, acceleration, throttle pedal actuation, and brake application) were used to evaluate the participants’ driving behavior, primarily on the basis of data from before and after the HVC installation. The before–after analysis allowed the assessment of HVC effectiveness in driver behavior modification. Mixed logit and random parameters linear regression models were estimated, and panel effects and unobserved heterogeneity were accounted for. Several factors were explored and controlled for (e.g., vehicle and driver characteristics, roadside environment, weather conditions), and the preliminary exploratory results show that HVCs can improve pedestrian safety and positively modify driving behavior.

A simulation approach to study emergency response

Simulation provides a significant tool in studying transportation systems and emergency response, allowing various scenarios of the ‘smart environments’ to be tested before real-world implementation. The simulation environment built in this paper uses Rockwell ARENA simulation software to provide a test bed for studying emergency response in a transportation network. With the data sources and assumptions used in building a roadway network with realistic traffic flow, accounting for both weather and congestion, a small study area in Western New York (WNY) provides the test bed based on real-life observations. After generating the expected movement of traffic, vehicular crashes are simulated, followed by Emergency Medical Service response. This paper focuses on the development process for building a simulation capable of modelling vehicular movement throughout a study area and a validation of the emergency vehicle travel times through historical crash response data in an existing traffic network. Its goal is to provide the basis of future work, enabling advanced transportation systems to be evaluated with respect to increased situation awareness resulting from optimized sensor placement, data fusion techniques and improved emergency response.

A Study of Situationally Aware Routing for Emergency Responders

This paper presents tests conducted on routes determined from a Dijkstra-based shortest path problem and a Variance-Constrained Shortest Path problem under varying conditions of traffic and weather in a simulated ‘smart environment’. Utilizing envisioned future advanced transportation systems’ real-time information on traffic parameters allows data fusion techniques to provide situation awareness to its users. Taking advantage of this real-time data, the routing methodologies and data capture techniques studied in this paper provides Emergency Medical Services with better routes when responding to a vehicular crash. Comparing the performance of both routing methodologies in terms of both their ability to provide better routes as well as computation times demonstrates two alternatives for aiding in future emergency response.

Locating Intelligent Sensors on a Transportation Network to Facilitate Emergency Response to Traffic Incidents

According to the Federal Highways Administration (FHWA) study presented in Office of Highway Policy Information there are 254 million registered motor vehicles in the U.S. Each year this number continues to grow, increasing utilization of the road transportation network. In 2009 alone there were an estimated 2.2 million injuries related to traffic incidents and 33,808 fatalities from these injuries. In addition, traffic incident related fatalities ranked sixth in the list of preventable fatalities in the U.S.

Improving the Quality of Weather Information Provided to Emergency Medical Services Responding to Motor Vehicle Crashes

When a motor vehicle crash happens, emergency medical services (EMS) offer the best prospects for injured occupants. It is therefore important to address any issue that can have a negative impact on the effectiveness of EMS response to a crash. One such issue relates to shortfalls in the accuracy and completeness of current and forecast weather information. These shortfalls are often attributable to limitations in the geographic resolution of measured data as well as difficulties with obtaining real-time access to updated weather as an event unfolds. This paper examines the need for improved weather information to support both ground and air EMS response. Operating environments for responders are examined, and established sources of measured weather data are described, in particular, the system of automated weather stations that support current and forecast weather reporting in the United States. Existing and emerging mobile sensor platforms, which could broaden the geographic extent of measured data especially as advanced intelligent transportation systems evolve, are then considered. Next, a review of weather-related issues identified in reports from helicopter EMS (HEMS) pilots and in severe weather after-action reports from municipalities is presented. For ground responders, a need for real-time, route-specific weather information was identified. For air responders, the system of airport-based weather observation stations developed for fixed-wing aircraft was found to be inadequate for HEMS. A rational, stepwise approach for expanding weather data collection to create a more spatially resolved, low-altitude weather information system to support HEMS is presented.