The study developed a new way to predict traffic crashes by looking at the safe distance between vehicles on highways. By using individual vehicle data, researchers found that this method was effective in predicting crash likelihood. They also discovered that a specific type of statistical model, called a hurdle negative binomial model, was better at predicting crashes than other models. This research shows that analyzing individual vehicle information can help improve freeway safety.