The Effects of Airbags and Seatbelts on Occupant Injury in Longitudinal Barrier Crashes

Introduction: Longitudinal barriers, such as guardrails, are designed to prevent a vehicle that leaves the roadway from impacting a more dangerous object while minimizing the risk of injury to the vehicle occupants. Current full-scale test procedures for these devices do not consider the effect of occupant restraints such as seatbelts and airbags. The purpose of this study was to determine the extent to which restraints are used or deployed in longitudinal barrier collisions and their subsequent effect on occupant injury. Methods: Binary logistic regression models were generated to predict occupant injury risk using data from the National Automotive Sampling System / Crashworthiness Data System from 1997 through 2007. Results: In tow-away longitudinal barrier crashes, airbag deployment rates were 70% for airbag-equipped vehicles. Compared with unbelted occupants without an airbag available, seat belt restrained occupants with an airbag available had a dramatically decreased risk of receiving a serious (MAIS 3+) injury (odds-ratio (OR)=0.03; 95% CI: 0.004- 0.24). A similar decrease was observed among those restrained by seat belts, but without an airbag available (OR=0.03; 95% CI: 0.001- 0.79). No significant differences in risk of serious injuries were observed between unbelted occupants with an airbag available compared with unbelted occupants without an airbag available (OR=0.53; 95% CI=0.10-2.68). Impact on Industry: This study refutes the perception in the roadside safety community that airbags rarely deploy in frontal barrier crashes, and suggests that current longitudinal barrier occupant risk criteria may over-estimate injury potential for restrained occupants involved in a longitudinal barrier crash.

Visualization Techniques For The Analysis Of Network Simulation Results

The Simulation Automation Framework for Experiments (SAFE) streamlines the de- sign and execution of experiments with the ns-3 network simulator. SAFE ensures that best practices are followed throughout the workflow a network simulation study, guaranteeing that results are both credible and reproducible by third parties. Data analysis is a crucial part of this workflow, where mistakes are often made. Even when appearing in highly regarded venues, scientific graphics in numerous network simulation publications fail to include graphic titles, units, legends, and confidence intervals. After studying the literature in network simulation methodology and in- formation graphics visualization, I developed a visualization component for SAFE to help users avoid these errors in their scientific workflow. The functionality of this new component includes support for interactive visualization through a web-based interface and for the generation of high-quality, static plots that can be included in publications. The overarching goal of my contribution is to help users create graphics that follow best practices in visualization and thereby succeed in conveying the right information about simulation results.