The problem of dilemma zone protection and red-light-running is especially important in certain rural intersections due to the higher speeds at these intersections and their isolated nature. In addition, the presence of a larger percentage of trucks mean that adequate warning and help need to be given to these truck drivers in order to enable them to stop safely, or proceed through the intersection before the onset of red. To curb any potential danger at such intersections, a Texas Department of Transportation (TxDOT) research project on Advanced Warning for End of Green Phase (AWEGS) at high speed intersections deployed AWEGS at two rural intersection sites ?? Tx 6 @ FM 185 near Waco and US 290 @ FM 577 in Brenham. The deployment of AWEGS involved a Level 1 and a later upgrade to a more efficient Level 2 in Waco. Initial results on red-light-running, even though promising, were expressed as observed red-light-running events per day. These resulting rates did not reflect exposure, and the results also raised some concerns with regards to some increase in red-light-running from Level 1 to Level 2. A more detailed analysis of the red-light-running issue at these two sites is provided in this thesis. The main areas of red-light-running analyses presented here are with respect to the reductions in red-light-running rates for the exposure factors of number of cycles and vehicular volumes, the comparison of day and night RLR rates and the nature of speeds of vehicles running the red light at the intersection in Waco. AWEGS was found to reduce the total red-light-running per exposure factor after its deployment. Both Level 1 and Level 2 AWEGS operations were found to reduce red-light-running by up to 60%. Generally, total red-light-running per exposure factor between Level 1 and Level 2 was found to be about the same. Level 2 had lower daytime red-light-running rates and higher nighttime rates than Level 1. Generally, day rates were found to be higher than night rates for all levels of AWEGS deployment. It is recommended that, to better understand the operational aspects of AWEGS and to improve its operations, more implementation of AWEGS and further tests be done. An automated method to collect and analyze data needs to be developed as well as a means of automatically recording video data for calibration and verification It is also recommended that Level 1 technology be implemented in areas where the Level 2 technology may be either too complex or too expensive.

Red light running is one of the main contributing factors of crashes in urban areas in Florida and the United States. Nationwide, according to preliminary estimates by the Federal Highway Administration (FHWA) 2001, there were nearly 218,000 red-light running crashes at intersections. These crashes resulted in as many as 181,000 injuries and 880 fatalities, and an economic loss estimated at $14 billion per year nationwide, According to the Community Traffic Safety Team Florida Coalition (A statewide traffic safety group) there were 9,348 crashes involving red-light running in Florida and 127 fatalities in 1999. This research study focused on studying the red light running crashes and violations in the State of Florida. There were three primary objectives for this research. The first primary objective was to analyze the red light running crashes in Florida from 2002 to 2004. The data for this part was collected from the Crash Analysis Reporting System of the Florida Department of Transportation. These crashes are reported as "disregarded traffic signal" as far as the first contributing cause. The analysis focused on the influences of different factors on red light running crashes including the driver (age group, gender, and DUI history) and the environment (time of day, day of week, type of road, and weather). However, not all red light crashes are reported as "disregarded traffic signal". Therefore, representing red light running crashes only through "disregard traffic signal" noted reports would underestimate the extent of red light running effects at a given intersection. Therefore, the second objective was to review the long form crash reports to determine the actual number of crashes related to red light running. The analysis for a random sample of the crashes on the sate roads of Florida on the year 2004 showed that the percentage of crashes related to red light running reported on the database was found to be (3.13%), and the percentage of crashes related to red light running reported in the original crash repot filled by the police officer are much higher than reported(5.63%), which shows the importance of standardizing the format and coding process for the long form crashes conducted by the police officers to help accurately identify the real cause of the crash at the studied location. The third objective was to analyze the violations data given for five intersections and find if there is a correlation between the average rate of violations per hour and the frequency of red light running crashes. The analysis showed that utilizing the limited number of intersections used in the study, it appears that there is no correlation between the average violations per hour and the red light running crashes at the studied locations.

This research explores the red light running phenomena and offer a better understanding of the factors associated with it. The red light running is a type of traffic violation that can lead to angle crash and the most common counter measure is installing a red light running cameras. Red light running cameras some time can reduce the rates of red light running but because of the increased worry of the public towards crossing the intersection it can cause an increase in rear end crashes. Also the public opinion of the red light running cameras is that they are a revenue generator for the local counties and not a concern of public safety. Further more, they consider this type of enforcement as violation of privacy. There was two ways to collect the data needed for the research. One way is through a tripod cameras setup temporarily placed at the intersection. This setup can collect individual vehicles caught in the change phase with specific information about their reactions and conditions. This required extensive manual analysis for the recorded videos plus data could not be collected during adverse weather conditions. The second way was using traffic monitoring cameras permanently located at the site to collect red light running information and the simultaneous traffic conditions. This system offered more extensive information since the cameras monitor the traffic 24/7 collecting data directly. On the other hand this system lacked the ability to identify the circumstances associated with individual red light running incidents. The research team finally decided to use the two methods to study the red light running phenomena aiming to combine the benefits of the two systems. During the research the team conducted an experiment to test a red light running countermeasure in the field and evaluate the public reaction and usage of this countermeasure. The marking was previously tested in a driving simulator and proved to be successful in helping the drivers make better stop/go decisions thus reducing red light running rates without increasing the rear-end crashes. The experiment was divided into three phases; before marking installation called "before", after marking installation called "after", and following a media campaign designed to inform the public about the use of the marking the third phase called "after media" The behavior study that aimed at analyzing the motorist reactions toward the signal change interval identified factors which contributed to red light running. There important factors were: distance from the stop bar, speed of traffic, leading or following in the traffic, vehicle type. It was found that a driver is more likely to run red light following another vehicle in the intersection. Also the speeding vehicles can clear the intersection faster thus got less involved in red light running violations. The proposed "Signal Ahead" marking was found to have a very good potential as a red light running counter measure. The red light running rates in the test intersection dropped from 53 RLR/hr/1000veh for the "before" phase, to 24 RLR/hr/1000veh for the "after media" phase. The marking after media analysis period found that the marking can help the driver make stop/go decision as the dilemma zone decreased by 50 ft between the "before" and the "after media" periods. Analysis of the traffic condition associated with the red light running it revealed that relation between the traffic conditions and the red light running is non-linear, with some interactions between factors. The most important factors included in the model were: traffic volume, average speed of traffic, the percentage of green time, the percentage of heavy vehicles, the interaction between traffic volume and percentage of heavy vehicles. The most interesting finding was the interaction between the volume and the percent of heavy vehicles. As the volume increased the effect of the heavy vehicles reversed from reducing the red light running to increasing the red light. This finding may be attributed to the sight blocking that happens when a driver of a passenger car follows a larger heavy vehicle, and can be also explained by the potential frustration experienced by the motorist resulting from driving behind a bigger vehicle.

The onset of a yellow indication is typically associated with the risk of vehicle crashes resulting from dilemma-zone and red-light-running problems. Such risk of vehicle crashes is greater for high-speed signalized intersection approaches. The research presented in this dissertation develops statistical as well as behavioral frameworks for modeling driver behavior while approaching high-speed signalized intersection approaches at the onset of a yellow indication. The analysis in this dissertation utilizes two sources of data. The main source is a new dataset that was collected as part of this research effort during the summer of 2008. This experiment includes two instructed speeds; 72.4 km/h (45 mph) with 1727 approaching trials (687 running and 1040 stopping), and 88.5 km/h (55 mph) with 1727 approaching trials (625 running and 1102 stopping). The complementary source is an existing dataset that was collected earlier in the spring of 2005 on the Virginia Smart Road facility. This dataset includes a total of 1186 yellow approaching trials (441 running and 745 stopping). The adopted analysis approach comprises four major parts that fulfill the objectives of this dissertation. The first part is concerned with the characterization of different driver behavioral attributes, including driver yellow/red light running behavior, driver stop-run decisions, driver perception-reaction times (PRT), and driver deceleration levels. The characterization of these attributes involves analysis of variance (ANOVA) and frequency distribution analyses, as well as the calibration of statistical models. The second part of the dissertation introduces a novel approach for computing the clearance interval duration that explicitly accounts for the reliability of the design (probability that drivers do not encounter a dilemma zone). Lookup tables are developed to assist practitioners in the design of yellow timings that reflects the stochastic nature of driver PRT and deceleration levels. An extension of the proposed approach is presented that can be integrated with the IntelliDriveSM initiative. Furthermore, the third part of the dissertation develops an agent-based Bayesian statistics approach to capture the stochastic nature of the driver stop-run decision. The Bayesian model parameters are calibrated using the Markov Chain Monte Carlo (MCMC) slice procedure implemented within the MATLAB® software. In addition, two procedures for the Bayesian model application are illustrated; namely Cascaded regression and Cholesky decomposition. Both procedures are demonstrated to produce replications that are consistent with the Bayesian model realizations, and capture the parameter correlations without the need to store the set of parameter realizations. The proposed Bayesian approach is ideal for modeling multi-agent systems in which each agent has its own unique set of parameters. Finally, the fourth part of the dissertation introduces and validates a state-of-the-art behavioral modeling framework that can be used as a tool to simulate driver behavior after the onset of a yellow indication until he/she reaches the intersection stop line. The behavioral model is able to track dilemma zone drivers and update the information available to them every time step until they reach a final decision. It is anticipated that this behavioral model will be implemented in microscopic traffic simulation software to enhance the modeling of driver behavior as they approach signalized intersections. / Ph. D.

It is estimated that as many as 2.7 million crashes occur each year at intersections or are intersection related; resulting in over 8500 fatalities each year. These statistics have prompted government and corporate sponsored research into collision countermeasure systems that can enhance safety at intersections. Researchers are investigating technologies to provide an infrastructure-based or infrastructure-cooperative Intersection Decision Support (IDS) systems. Such systems would use pre-specified algorithms to identify drivers that have a high likelihood of violating the traffic signal and thus increase the risk of a collision. The system would subsequently warn the violating driver to stop though an in-vehicle or infrastructure-mounted interface. An IDS algorithm must be designed to provide adequate time for the driver to perceive, react, and stop the vehicle, while simultaneously avoiding a high false alarm rate. Prior to developing these algorithms, scientists must understand how drivers respond to traffic signals. Little research has focused on the influence of driver state on red-light running behavior or methods for distinguishing red light violators from non-violators. The objective of the present study was to define trends associated with intersection crossings under different driver states and to explore the point detection method of predicting red light running upstream of the intersection. This was accomplished through a test-track mixed-factor experiment with 28 participants. Each participant experienced a baseline (complete a full stop at the red light), distracted (misses signal phase change due to inattention), and willful (driver knowingly makes a late crossing in an attempt to 'beat the light') driver state conditions. To provide the opportunity for red-light running behavior from participants, the amber change interval began at five different distances from the intersection. These distances were located near and within the dilemma zone, a region in which drivers have a difficult time deciding whether to go or to stop. Data collected from in-vehicle sensors was statistically analyzed to determine significant effects between driver states, and to investigate point detection algorithms. / Master of Science

Crashes categorized as running red light or left turning are most likely to occur at signalized intersections and resulted in substantial severe injuries and property damages. This dissertation mainly focused on these two types of vehicle crashes and the research methodology involved several perspectives. To examine the overall characteristics of red-light running and left-turning crashes, firstly, this study applied 1999-2001 Florida traffic crash data to investigate the accident propensity of three aspects of risk factors related to traffic environments, driver characteristics, and vehicle types. A quasi-induced exposure concept and statistical techniques including classification tree model and multiple logistic regression were used to perform this analysis. Secondly, the UCF driving simulator was applied to test the effect of a proposed new pavement marking countermeasure which purpose is to reduce the red-light running rate at signalized intersections. The simulation experiment results showed that the total red-light running rate with marking is significantly lower than that without marking. Moreover, deceleration rate of stopping drivers with marking for the higher speed limit are significantly less than those without marking. These findings are encouraging and suggesting that the pavement marking may result in safety enhancement as far as right-angle and rear-end traffic crashes at signalized intersections. Thirdly, geometric models to compute sight distances of unprotected left-turns were developed for different signalized intersection configurations including a straight approach leading to a straight one, a straight approach leading to a curved one, and a curved approach leading to a curved one. The models and related analyses can be used to layout intersection design or evaluate the sight distance problem of an existing intersection configuration to ensure safe left-turn maneuvers by drivers.

Drivers' mistakes in making immediate decision facing yellow signal interval to stop or go through the intersection is one of main factors contributing to intersection's safety. Incorrect decision might lead to a red light running and a right-angle Collison when passing through the intersection or a rear-end collision when failing to stop safely.Improperly timed traffic signal intervals result in the inability of the drivers to make the right decision and can place them in the dilemma zone. Advance warning systems (AWS) have been used to provide information about the downstream traffic signal change prior to approaching the intersection. On the other hand, advance warning systems increase drivers approach speed according to the literature. However, effect of AWS on dilemma zone has not been studied before. The goal of this thesis is to minimize the number of vehicles caught in dilemma zone by determining more precise boundaries for dilemma zone and to reduce the number of red light violations by predicting the red light runners before arriving to the intersection. Here, dilemma zone boundaries at the presence of AWS has been reexamined with the aid of a large dataset (more than 1870 hours of data for two different intersections). Upper dilemma zone boundaries found to be higher for the intersections with AWS. This is due to vehicles' increasing the speed at the flashing yellow sings to escape the dilemma zone.Moreover, an algorithm for predicting red light runners and distinguishing them from right turners is presented. / Master of Science

Advance warning signals installed upstream of a high-speed signalized intersection (HSSI) warn motorists of impending signal changes in an effort to reduce the frequency of red-light running (RLR) and crashes. A new advance warning signal design was tested on an approach to an HSSI in Utah to study the effects of the modified design on motorist behavior. The new design utilized an overhead dynamic blank-out sign and flashers. A state-of-the-art digital wave radar evaluation system was installed at the study site to collect continuous data of vehicle speeds and RLR events by a non-intrusive method. Crash data were collected from the jurisdiction responsible for the study site and for an additional control intersection. Data were collected prior to, immediately after, and eight months after installation The blank-out overhead dynamic advance warning signal (BODAWS) system reduced RLR at the site during the time period immediately after installation. Eight months later, the number of RLR violations was slightly higher on one approach than before BODAWS system installation. Crash results showed that six months after BODAWS installation, the number of crashes declined at the study site. The number of crashes proportionately declined at the control intersection as well indicating a need to continue to evaluate and monitor changes. Mean vehicle speeds recorded before the onset of the yellow signal increased on the approaches to the study site immediately after BODAWS installation, and remained higher eight months later. Mean vehicle speeds recorded during the yellow signal, increased eight months after BODAWS installation to speeds higher than before the system was installed. Higher speeds during the yellow signal, combined with an increase in the number of RLR violations eight months after BODAWS installation, suggest that motorists may have begun to use the advance warning to speed up in an attempt to enter the intersection before the signal turned red. It is recommended that the lead flash time between activation of the BODAWS signs and flashers and the onset of the yellow signal should be adjusted so that motorists are not provided with more time than is necessary to safely clear the intersection.