A Comparison of Ambulance Travel Time Approximation: Using Google Maps and Machine Learning

Abstract

Accurately estimating ambulance travel time is critical for emergency medical services (EMS) to ensure timely response and efficient resource allocation. This study explores machine learning-based approaches for predicting ambulance travel time, comparing three artificial neural networks (ANN), Google Maps, and the traditional KWH constant speed model. Using real-world data from Nova Scotia's EMS system, we develop and evaluate these models based on prediction accuracy, computational efficiency, and applicability in both real-time and modeling contexts. Our results show that ANN model aligns closely with historical data, making it more suitable for EMS simulation and decision-making, despite requiring more complex feature inputs. This study highlights the strengths and limitations of each method and offers insights into selecting the most appropriate approach based on EMS application requirements.