By Robbie Webber
Researchers at Brigham Young University have developed one of the most advanced traffic safety models in the country, and Utah DOT will be using it to target allocation of scarce funds to the most crucial locations to save lives and prevent serious injuries.
Civil engineering professor Grant Schultz and two colleagues—civil engineering professor Mitsuru Saito and statistics professor Shane Reese—used Bayesian statistics to build a model that will predict “hot spots” where more crashes than normal are likely to occur.
The model uses more variables than just crash numbers and traffic volume to get a clearer picture of the problem. It also takes into account all of the conditions of a segment of roadway and determines how many crashes are expected to occur there, Shultz said, then it compares that number with the actual number of crashes that happen.
The technique, he said, is a game-changing way to account for many sources of uncertainty in a problem — like individual drivers, road conditions, and the number of curves in a stretch of highway.
It took five years to perfect the Utah Crash Prediction Model, but UDOT and the Utah Department of Public Safety have been so pleased with the early results that they recently awarded the trio of professors the Executive Director’s Excellence in Transportation Safety Award.
Robbie Webber is a Senior Associate at SSTI.