Event Details

Sebastian Birolini of University of Bergamo will present a seminar to the Management Science Department.

Abstract: Flight delays are significant sources of disruption and unexpected costs in airline operations. Gaining early and accurate visibility into delays is crucial for building robust flight sequences that can absorb delays and minimize their propagation. However, delays are difficult to predict in advance, further complicated by the fact that historical data includes both primary delays (arising from external factors) and propagated delays (resulting from cascading effects within the airline’s network). Under this premise, this research proposes predictive and prescriptive analytics models to forecast primary delays and optimize day-ahead aircraft routing for proactive delay mitigation. First, we develop a quantile regression model to disentangle primary delays from historical data, and an ensemble machine learning model to predict delays based on flight-level, environmental, and traffic features—estimated using a queuing model of airport operations. We then integrate our data-driven delay predictions into deterministic and stochastic aircraft routing models to support day-ahead planning. Using real-world data from Vueling Airlines, we evaluate the models out-of-sample against real-world counterfactuals. Results show that our predictive model achieves a mean absolute error of 7–8 minutes, and that our prescriptive models can reduce delay costs by 3–5%, practically underscoring the benefits of advanced predictive and prescriptive analytics to enhance the robustness of airline operations.