October 23, 2014

About the Presentation

This seminar explored two related questions. First, can investigation and reconstruction of individual road crashes be used to estimate the effects of safety-related countermeasures? Conditions were given under which prior background knowledge plus consistent estimation of the conditions initiating crashes lead to a lower bound for a crash modification factor. If the countermeasure’s effect is monotonic, this bound becomes tight.

Second, since estimation in crash reconstruction is model-based, how can model assumptions be checked and validated? In statistical modeling, cross-validation refers to the practice of fitting a model with part of the available data and then using predictions of the unused data to identify possible weaknesses in the model. In crash reconstruction, a partial cross-validation is possible when two different measurements can be used to estimate the same crash feature. An implementation of this approach was described. Throughout, main ideas were illustrated using simulated, staged, and actual vehicle/pedestrian crashes.

About the Speaker

Gary Davis

Gary Davis is a professor of civil, environmental, and geo- engineering at the University of Minnesota. His research interests include causal inference and impact assessment in traffic safety, application of accident investigation and reconstruction methods to traffic engineering questions, the use of Bayesian statistical methods in traffic and transportation engineering, and the application of optimization methods to problems in traffic engineering and transportation planning.