Jiguang Zhao

Safety performance evaluation of left-side off-ramps at freeway diverge areas

The safety performance of left-side off-ramps was evaluated by comparing that of right-side off-ramps at freeway diverge areas. Crash records at a total of 11 left-side and 63 similar right-side diverge areas in Florida were collected. Based on the data collected, the traffic conflict study and the cross-sectional comparison of crashes were conducted in this study. Four types of traffic conflicts were identified and counted. The average conflict rates near the ramp area were found to be approximately 10 per 1000 conflicting vehicles. Crash data were compared for the left-side off-ramps with right-side off-ramps by two exit ramp types: one-lane exit and two-lane exit with an optional lane, respectively. The comparisons indicate that the left-side off-ramp did have higher average crash counts, crash rate and percentage of severe crashes, but the difference is only statistically significant for the severe crashes at a 10% level. A crash prediction model for one-lane exit was developed to identify the factors that contribute to the crashes that have been reported for selected freeway segments. The conclusion is consistent with cross-sectional comparison. It is expected that this study could help engineers have a better understanding of left-side off-ramps at freeway diverge area and select the appropriate countermeasures and practical designs.

Safety Effects of Median Treatments Using Longitudinal Channelizers: Empirical Bayesian Before-and-After Study

AbstractA new median treatment technique using longitudinal channelizers has been used in Florida as an alternative to the traditional permanent concrete barrier. However, the effects of the new technique on roadway safety and its economic performance have not been well documented. This paper’s intent is to present the results of a comprehensive evaluation on the safety and economic performance of the longitudinal channelizers’ median treatments with before-and-after crash data. Two types of longitudinal channelizers’ treatments—full median opening to directional median opening and full median opening to left-in only median opening—were evaluated. To account for the site selection bias, the empirical Bayesian before-and-after study method was used. The safety performance function for roadway segments with full median openings was developed with the generalized linear model to consider the negative binomial distribution of the crash counts in population entities. The targeted crash types for the longitudin...

Correlating the Safety Performance of Urban Arterials with Lighting: Empirical Model

This paper presents a new method to evaluate roadway lighting performance at the corridor level based on the illuminance data collected by a mobile lighting measurement system. The illuminance data were collected at 40 ft (12 m) intervals on the outside and inside lanes for both directions of the roadway. Statistical analysis results indicate that the illuminance data at the corridor level fit the lognormal distribution. Two parameters, the mean and the variance of the illuminance, could be used to evaluate the lighting system at the corridor level by comparing the parameters for the roadway lighting systems with those for the design standard. In addition, the daytime and nighttime crash rates for the study corridors were calculated. The results indicate that the illuminance parameters are significantly related with the difference between daytime and nighttime crash rates; this factor further justifies the reasonableness of the new lighting evaluation method. The proposed method is recommended for evaluating lighting systems at the corridor level when the data are collected with the new mobile lighting measurement system.

Impact of Safe Routes to School on Student Travel Mode Shifting: Based on the Before and After Period Survey

An evaluation of the effectiveness of the Safe Routes to School (SRTS) program was conducted with data collected before and after the implementation of the SRTS improvements at thirteen schools in Hillsborough County, Florida. Descriptive Statistics indicate that the students’ walking/biking rates and participants’ subjective feelings toward walking/biking activity improved significantly at some schools. Those schools with significant increase in walking/biking rates have similar characteristics. The school-level walking/biking rates and associated factors are discussed. A linear regression model is established to estimate the school-level walking/biking rates. Students’ walking/biking rates is a function of weighted distance to school, weighted grade level, male student percentage, and average student family size. Other factors not included in the model, such as the walking/biking environments in school areas, are found to play an important role as well. Model calibration results indicate a school flasher sign alone could increase the school-level walking/biking rates by about 3%.

Identify Factors Affecting Number of Students Walking or Biking to School

This paper reports on a study that was conducted to investigate the characteristics of student travel behaviors before the implementation of a Safe Routes to School (SRTS) program. It also seeks to identify the influential factors that affect the number of children who walk or bike to school. As part of the Florida SRTS program, a pilot survey was conducted for both students and parents in Hillsborough County, Florida before the program started. Student survey results showed that student demographic characteristics such as grade, gender, the number of children in the family, and the distance from home to school influence their travel mode to school. Parents subjective opinions were also significantly associated with students travel mode to school. Distance, intersection safety and weather/climate were considered affective on a parents decision on whether to allow their child to walk or bike to school.