Surrogate measures for evaluating new signage and intersection designs

Abstract

Transportation agencies faced with the challenge of enhancing safety on roadways are looking for alternative solutions to designing roads and signage. When deciding whether the alternative design is superior to the traditional one, decision makers need methods and quantitative data to evaluate these alternatives. This dissertation provides two accessible methods to compare different alternative designs and illustrates them using case studies. The first method involves using speed-based statistical measures that are extracted from video-based traffic surveillance. This method was more accurate in collecting vehicle speeds than the speeds extracted from video-based data collection systems. It was then utilized to evaluate the effectiveness of an alternative merge sign in work zones. This alternative sign consists of an arrow pointing the merge direction and text describing the lane closure, while MUTCD sign is graphical. The case study measured driver behavior characteristics including speeds and open lane occupancies. The results indicate that open lane occupancy was higher for the test sign in comparison to the MUTCD sign upstream of the merge sign. The occupancy values at different distances between the merge sign and the taper were similar for both the test and MUTCD signs, but the test sign encouraged up to 11% more cars to be in the open lane immediately upstream of the merge sign. Passenger cars stayed in the closed lane longer, or closer to the taper, than trucks. The merging behavior of truck drivers did not vary significantly with the type of merge sign deployed in the work zone. The analysis of speed characteristics did not reveal substantial differences between the two sign configurations. The mean speeds with the MUTCD configuration were 1.3 mph and 2 mph lower than the test configuration at the merge sign and taper locations, respectively. The second method utilizes microscopic traffic simulation to evaluate alternative designs. This method is ideal for projects where video monitoring of the entire study of interest is not feasible. Evaluating alternative designs with crash data usually requires a long time span to build the facility and record crash data over at least one year after the facility has been open to traffic. In addition to that, new facility needs to be built or altered if other design features are to be tested. With microscopic simulation, the time cost for the study is greatly shortened and different design aspects can be tested in a risk-free environment. Two case studies are presented to illustrate this simulation method. The first case study involves a work zone while the second case study focuses on evaluating a J-turn intersection design. The spacing of U-turn and the inclusion of acceleration and deceleration lanes were evaluated, in the J-turn study. A simulation analysis was conducted to study the impact of different design variables on the safety of J-turns. A base simulation model was created and calibrated using field data collected in a previous Missouri Department of Transportation (MoDOT) project on J-turns. The calibrated model was then used to study various combinations of major road and minor road volumes and design variables. The simulation analysis helped develop guidance on recommended spacing for various major road and minor road volume scenarios. For all the studied scenarios, the presence of acceleration lane resulted in significantly fewer conflicts. Thus, acceleration lanes are recommended for all J-turn designs, including lower volume sites. Second, while U-turn spacing between 1000 feet and 2000 feet was found to be sufficient for low volume combinations, a spacing of at least 1500 feet and 2000 feet are recommended for medium and high volumes, respectively.