Predictive models for urgency perception in railway crossing alarm signals: Development and applications for Argentina

Abstract

This study develops predictive models for the perceived urgency of railway crossing alarm signals, focusing on both frequency and temporal variables. Two subjective tests using incomplete pairwise comparison matrices are conducted to evaluate the effects of these variables. In the first test, pitch and harmonic attenuation are examined as frequency-related factors, while in the second test, speed and attack time are assessed as temporal factors. For each test, nine synthetic alarm signals are generated based on the alarm sounds used in Buenos Aires. The results demonstrate that all four variables significantly influence perceived urgency, with temporal variables exerting a stronger effect than frequency variables. Multiple regression analyses reveal that higher harmonic amplitudes and increased pitch enhance perceived urgency, with harmonic attenuation playing a more dominant role. Similarly, faster speeds and shorter attack times heighten perceived urgency, with nearly equal contributions from both variables. Based on these findings, the study proposes optimized alarm configurations to maximize perceived urgency while adhering to Argentinean regulations. These insights highlight opportunities for improved alarm design to enhance safety at railway crossings.