Modelling driver behaviour towards innovative warning devices at railway level crossings

Area of Science:

• Transportation Engineering
• Human Factors in Driving
• Traffic Safety Analysis

Background:

• Railway level crossing safety is a significant concern, often limited by funding for improvements.
• The impact of various warning systems on driver behavior requires further investigation for cost-effective solutions.

Purpose of the Study:

• To compare driver behavior in response to novel (rumble strips, in-vehicle audio) and conventional (flashing light, stop sign) railway level crossing warning devices.
• To develop regression models predicting driver responses and stopping probabilities.

Main Methods:

• A driving simulation experiment was conducted to collect driver response data.
• Binary choice and mixed regression models were developed to analyze driver behavior, including stopping probability and timing of actions.
• Factors such as age, gender, and speed were analyzed for their influence on responses.

Main Results:

• Active warning systems (rumble strips, in-vehicle audio) demonstrated significantly higher driver compliance than passive systems (flashing light, stop sign).
• Driver behavior varied significantly based on age, gender, speed, and the type of warning device encountered.
• Regression models successfully predicted driver responses at different stages of approach to railway crossings.

Conclusions:

• Active warning systems offer a more effective approach to improving driver compliance at railway level crossings.
• Behavioral models and simulation tools can be utilized to estimate traffic safety indicators and compare the relative safety of different warning devices.