Click the title of the project to see the full final report.
Click on the image to read a summary about the project (not available for all projects).
This research project explored the content of SHRP2 NDS dataset and performed statistical analysis to identify useful performance measures to detect distracted driving behavior as well as provided an outline for a crash index model that can be used to quantify the crash risk associated with distracted driving behavior. The objective of the research was to detect whether a driver was engaged or not in one of three specific secondary tasks (talking or listening on a hand-held phone, texting/dialing on a hand-held phone, and driver interaction with an adjacent passenger) using selected performance measures. Although multiple logistic regression analysis did not provide a statistically good fit of the data, the neural network analysis proved to be good tool for detecting drivers’ engagement in secondary tasks. The research team proposed a framework of crash index calculation that can be used to quantify the crash risk associated with distracted driving behavior.
The overall goal of this project was to identify factors that affect seatbelt use in Louisiana which can be used to develop strategies leading to a significant increase in compliance rates. This project concentrated on the group of unbelted occupants to determine additional factors that can be used for effective strategies to increase seatbelt use in the state. The existing nature of seat belt use and enforcement in Louisiana was examined using secondary data analysis of citation data, roadside survey data, and crash data. To better
understand why some people do not use seat belts 100% of the time, primary data was collected via survey research methods. Findings indicate that despite a large number of citations being issued many drivers continue not to use seat belts; however, these individuals are not easily categorized by traditional demographic factors. Rather than categorizing individuals as “users” and “non-users,” people tend to fall into one of four groups of belt users: motivated commitment, behaviorally compliant, well-intentioned, and generally disinclined. Consistent belt use depends heavily on motivation, habit, and routine. A multistate sample comparing Louisiana drivers with drivers in other states suggests individuals with less-than-perfect belt use are similar across states.
The objectives of this research were to evaluate laws and policies about drugged driving in Louisiana and other states, identify obstacles to a per se and ZT law for drugged driving, assess the availability of drugged driving data, and provide an analysis of the frequency of drugged driving to identify ways to improve data collection on drugged driving. Findings reflect a lack of standardized procedures and an uneven distribution of resources throughout Louisiana. A comparison of DWI arrests, speeding violation and crashes of drivers who tested positive for various drugs shows that drivers arrested for drugged driving have higher rates of prior unsafe driving incidents than other drivers. Survey interviews reveal an overall lack of training, resources, and testing capacities in Louisiana. This study gives a clearer understanding of existing data limitations and challenges, and presents recommendations for dealing with drug-impaired driving.
This study developed an AADT estimation methodology by using modern statistical and pattern recognition methods. By using available traffic counts on non-state roadway and four variables (namely: population, job, and distance to intersection and to major state highways at block level), a training set to estimate roadway AADT for eight parishes was obtained by a modified support vector regression method. This pattern recognition method yielded better AADT estimates than the conventional parametric statistical methods. Sensitivity analyses were also conducted in this study, which indicated a parish-specific model worked better than an aggregated single model.
The principal objectives and scope of this project were to provide a software tracking tool to improve decision-making for highway safety. A literature search revealed that purchasing and customizing existing software was not feasible and a new solution would be developed in-house. Requirement gathering and analysis was conducted and documented. The application was programmed as a web-based solution for collecting data on low-cost safety improvements and analyzing the effectiveness of the improvements. All programming and testing was conducted in house. The application was piloted by the Louisiana Department of Transportation and Development (DOTD).
This project developed a teaching package for safety fundamentals for undergraduate students and graduate students in civil engineering. The course covers seven topics: introduction to highway safety, basic safety concepts, safety-related data, fundamental statistics, development of safety models, safety predictive models in HSM, and safety evaluation. Accordingly, seven lecture notes were developed along with homework assignments, quizzes, and exams.
The goal of this project was to develop Louisiana state-specific HSM calibration factors for eight facility types. Some of the resulting factors were unexpected, in particular, those for urban three lane and urban five lane highways which were lower than anticipated. The calibration factors for rural two-lane, rural multilane undivided and divided, urban/suburban two lane, and urban/suburban four-lane divided and undivided highways, ranged from a low of 0.62 for rural multilane undivided highways to a high of 2.54 for urban/suburban four lane divided highways. It is expected that with an understanding of the conditions under which these factors were developed, they will be acceptable for use by analysts seeking to conduct highway safety analyses for roads in Louisiana.
The objective of this research was to use the driving simulator on the LSU campus to measure the risks associated with various distractions faced by the driving population. Participants were placed in simulated environments while being exposed to different driver distractions (handheld cell phone conversation, texting, and front-seat passenger conversation) to determine the effect on the driving task. The study concluded that texting and talking to passengers while driving impaired driver performance, but it did not find significant effects for cell phone conversation.