A comparative analysis of freeway crash incident clearance time using random parameter and latent class hazard-based...

Area of Science:

• Transportation Engineering
• Econometrics
• Traffic Management

Background:

• Freeway incident clearance times significantly impact traffic flow and have substantial economic consequences.
• Understanding factors influencing incident duration is critical for effective traffic management.

Purpose of the Study:

• To investigate factors affecting freeway incident clearance times in Alabama.
• To compare the performance of random parameters duration modeling and latent class duration modeling in explaining incident durations.
• To identify the most suitable econometric model for analyzing unobserved heterogeneity in incident clearance times.

Main Methods:

• Utilized a unique dataset of 2206 freeway crash incidents from Alabama in 2018.
• Applied advanced econometric methods: random parameters duration modeling and latent class duration modeling.
• Integrated data from multiple sources including ALDOT, CAPS, ASAP, and HPMS.

Main Results:

• The random parameters model identified 19 statistically significant variables (5 random, 14 fixed).
• The latent class model identified 18 statistically significant variables across two distinct classes (23% and 77% probability).
• The latent class model demonstrated a superior fit for the incident duration data.

Conclusions:

• The latent class duration model is more effective in accounting for unobserved heterogeneity in freeway incident clearance times.
• Findings provide data-driven evidence to inform policy and reduce incident clearance durations.
• This research contributes to the body of knowledge on traffic incident management and econometrics.