Incident Response System to Assist Active Traffic Management in D.C.

Incident Response Control Center

Improving incident response through better utilization of data can help save travel impact and lives.

This project aims at improving incident response strategies by exploring historical incident and traffic data. Traffic incidents have become a major cause of congestion and significant threat to urban mobility. Many road networks in major cities are currently operating near, if not beyond, capacity during peak hours. Capacity reduction and road closure due to incidents can cause significant delays over an extended period. An effective incident management system not only helps to mitigate congestion through swift incident detection, response, and site clearance, but also generates significant environmental benefits by reducing fuel consumption, emissions, and potential secondary incidents.

By exploring both historical incident and traffic data, the system can be improved by proactively deploying response units. The system should adapt itself to evolving incident patterns over different time of day and under different traffic/weather conditions, and change the strategy accordingly. Moreover, an effective system must also consider the network effect and travel behavior in response to changed traffic conditions in the aftermath of major incidents. These factors are extremely important in an urban setting where traveler information system is usually readily available and multiple alternative routes co-exist. This study would address these challenges.

There have been many studies on traffic incident management. Most incident response strategies proposed in the literature can be classified into two categories: dispatching and patrolling systems. Some researchers (Skabardonis et al. 1998; Lou et al. 2010) investigated the freeway service patrols (FSP) program under which incident response units will constantly roam on freeways to detect and respond to traffic incidents. The key question for such a system is how to divide a traffic network into independent patrol segments and how to assign response units to them. In contrast, Larson and Odoni (1981) and Pal and Bose (2009) argued that it is more efficient to deploy response units strategically and dispatch them after an incident has been detected. In previous studies, Zhu et al (2013, 2014) investigated both systems using data provided by the Coordinated Highways Action Response Team (CHART) in Maryland. They found that both paradigms of response strategies have their advantages in practice, depending on the incident locations and some operational constraints. The optimal design for incident response strategy varies over different time periods (peak versus nonpeak) and thus with different incident/traffic patterns. Therefore, it could be beneficial if the response strategy could adapt itself to the varying conditions.

Designing such an adaptive incident response system would require a better understanding of the incident/traffic patterns. As it takes time and resources to change the deployment, the system must learn from historical data and have predication capability to a certain level in order to make it functional. With the development and deployment of active traffic management strategies in more and more metropolitan areas, more real-time/near real-time data would become available. These data would further create new opportunities to enhance such an adaptive incident response system. Successful deployment of such system could in turn improve corridor management. Northern Virginia area suffers from severe traffic congestion due to high travel demand and frequent traffic incidents. This study would take this area as an example to evaluate the efficiency of an adaptive incident response system.

Task 1: Exploring historical traffic and incident data in Northern Virginia area to capture prevailing traffic/incident patterns during different time of day, and potentially, under different weather conditions. This effort would also help to inform VDOT what a “typical” traffic pattern looks like in the area.

Task 2: Developing a model to estimate/predict incident patterns (or lack of patterns) during different time of day and under different traffic conditions. Outputs from this model would inform the development of incident response strategies and allow potential pre-actively deployment of resources.

Task 3:  Developing an optimized pre-positioning/dispatching incident response system using data collected from the Northern Virginia area. The research team will select the corridor/area for a case study in consultation with VDOT regional district office.

Task 4: Developing an optimized patrolling strategy using the data collected from the Northern Virginia area.

Task 5: Comparing the performance of different response strategies and exploring the possibility to develop a flexible and adaptive response system.

Task 6: Exploring the possibility and methodology to consider real-time traffic information and behavioral changes under incident conditions.

Researchers:

·         Principal Investigator: Shanjiang Zhu, George Mason University

·         Co- Principal Investigator: Mohan Venigalla  GMU

Partners: Virginia Department of Transportation (VDOT) and Maryland Department of Transportation (MdDOT)

Data Sources: (1) Historical traffic incident data and network performance data from VDOT; (2) Historical traffic incident data and traffic response system performance data from the Coordinated Highways Action Response Team (CHART) in Maryland.