Drivers’ actions in an intersection’s dilemma zone – the area where the decision to stop at a yellow light or continue through it is not clear-cut – can lead to side-angle and rear-end crashes. In Maryland, researchers developed an intelligent dilemma zone protection system (DZPS) that is reducing these crashes by anticipating drivers’ decisions and responding. The DZPS system was deployed at two high-speed rural intersections (US 40@Western Maryland Parkway and MD 213@Williams/Locust Point Road), and it has three components:
(1) two wide-range sensors to track the speeds and locations of all vehicles within the identified dilemma zones;
(2) software to predict the response of drivers during the yellow phase and to activate the all-red extension function if needed; and
(3) a web-based module for responsible engineers/technicians to monitor the system’s performance from a control center.
Measured benefits of DZPS include a 30 to 40 percent reduction in dilemma zone length and fewer vehicles approaching the intersection at speeds greater than the posted speed limit. The all-red extensions have helped prevent crashes between through traffic and vehicles entering the intersection from the cross street.

Download (MD-17-SHA-UM-3-32_DMZ_Report.pdf)

Authors: Mark L. Franz and Gang-Len Chang
Date: July 21, 2014
Abstract:
This research project developed vital operational guidelines for design of a variable speed limit (VSL) system and its integrated operations with ramp metering control in contending with recurrent highway congestion. The developed guidelines can serve as an effective tool for traffic engineers to determine when to activate a VSL control and under what traffic conditions it needs to be supplemented by ramp metering operations to ensure the stability of traffic evolution over the congested highway segment. This report also presents various measures of effectiveness for evaluating the benefits of VSL and its integration with ramp metering control. A VSL control algorithm to compute the time-varying speeds in real time, based on detected traffic conditions, was developed in this study. Extensive simulation experiments, calibrated with the field data from US100 in Maryland, were conducted to evaluate the effectiveness of the developed VSL algorithm. Both the experimental results and sensitivity analyses with respect to key model parameters confirmed that proper implementation of VSL can indeed mitigate the congestion caused by the high-speed variance among vehicles and allow traffic flows to better utilize the available roadway capacity.

Download (UMD-2012-03.pdf)

Authors: Dr. Gang Len Chang, Dr. Cinzia Cirillo, Nayel R. J. Urena Serulle & Pratt Hetrakul
Date: Dec. 2011
Abstract:
Road pricing has been advocated as an efficient travel demand management to alleviate congestion since the seminal work by Pigou (1920) and Knight (1924) (see Lindsey, 2006, for recent reviews). More specifically, dynamic toll pricing has received greater interest among policy makers and public agencies due to its potential for lowering energy costs for society. Some analytical studies (e.g., Arnott et al., 1990) have found that dynamic toll pricing generally yield greater efficiency gains than static toll pricing because the former reduce queueing delays by altering travelers’ departure times as well as routes.
The construction of the Inter-county Connector (ICC) has certainly offered the prospect of reducing travel time between the I-270 and I-95 corridors, and may potentially alleviate congestion on the I-270 and I-495. Given that the ICC relies on dynamic toll pricing scheme, its daily traffic volumes are governed by individual trip-makers’ perceived time and cost saving in the term of value of travel time (VOT). Moreover, the ability to realistically capture trip-makers’ responses to time-varying road charges in term of willingness to pay (WTP) for toll is essential for predicting network flows and network equilibrium assignment models. These behavioral characteristics of users vary across individuals. Therefore capturing the heterogeneity of users in this regard is critical in predicting the impact of dynamic pricing schemes (e.g., Lu et al., 2008).
This study proposes the model that enables practitioners to integrate user response to dynamic toll pricing. The analysis accounts for cost and time savings perceived by regional drivers and the users’ response to time-varying road charges. More specifically, the study captures difference in behavioral characteristics of the willingness to pay (WTP) for toll across users socioeconomics and trip related characteristics such as time of day,

Download (UMD-2009-09.pdf)

Authors: Gang-Len Chang & Zichuan Li
Date: Nov. 2010

Download (UMD-2008-03.pdf)

Authors: Gang-Len Chang & Deepak Shrestha
Date: May.2000

Download (PERFORMANCE-EVALUATION-OF-CHART-AN-INCIDENT-MANAGEMENT-PROGRAM-IN-1997.pdf)

Author: Gang-Len Chang
Date: 2013-2014

Download (NTC-2014-Information-Form_Principal-Investigator_Gang-Len-Chang_September-2014.pdf)

Authors: Jie Yu & Yang Lu
Date: Nov. 14, 2008

Download (Methods-for-Identifying-hazardous-Locations-A-Comparison-study.pptx)

Authors:  GANG-LEN CHANG, MARK L. FRANZ & YANG LU
Date: Aug. 2012
Abstract:
The goal of this research was to develop a system that utilizes the dynamic detection technology and to evaluate the performance of this system. The first step of this study was to select an intersection experiencing a high frequency of crashes that could be remedied by a dilemma-zone protection system. Next, a dynamic dilemma-zone protection system was designed using three microwave sensors to track vehicles approaching the intersection on the major approach. The data collected by these sensors were then used in real-time to control the signal logic, providing green extensions and all-red extensions when pre-defined parameters of detected vehicles are met. To evaluate the performance of the system design and the appropriateness of the associated parameters, a field test was conducted. The data analysis included the identification of false called all-red extensions (related to efficiency) and missed all-red extensions (related to safety) to assess the overall performance of the newly installed dynamic dilemma zone protection system.

Download (MD-12-SP708B4J_Dilemma_Zone_Warning_System_Final_Report.pdf)