Predicting on-time delivery in the trucking industry

Abstract

On-time delivery is a key metric in the trucking segment of the transportation industry. If on-time delivery can be predicted, more effective resource allocation can be achieved. This research focuses on building a predictive analytics model, specifically logistic regression, given a historical dataset. The model, developed using six explanatory variables with statistical significance, results in a 76.4% resource reduction while incurring an impactful error of 2.4%. Interpretability and application of the logistic regression model can deliver value in predictive power across many industries. Resulting cost reductions lead to strategic competitive positioning among firms employing predictive analytics techniques.