Assessment of traffic safety at unsignalized intersections using post encroachment time under heterogeneous traffic conditions

Abstract Road safety assessment at unsignalized intersections is particularly challenging in developing countries due to the limited availability and reliability of crash data. Surrogate safety measures (SSMs) provide an effective alternative by enabling proactive evaluation of traffic conflicts without relying on historical crash records. Among these indicators, Post-Encroachment Time (PET) has been widely used to quantify the temporal separation between conflicting road users and to identify potentially hazardous interactions. This study evaluates the safety performance of six unsignalized intersections in Patna, India, using PET as the primary surrogate safety indicator. Traffic interactions were recorded using high-mounted video cameras during peak hours, and vehicle speeds, volumes, and conflict events were extracted through frame-by-frame analysis using Kinovea software. Manually derived PET values were compared with PET estimates obtained from microsimulation using the Verkehr In Städten–SIMulationsmodell (VISSIM) and the Surrogate Safety Assessment Model (SSAM). The results indicate significant variation in conflict frequency across intersections. Critical conflicts accounted for approximately 25% and 48% of total interactions at Beur More and Bypass Road, respectively. The comparison further revealed that manual PET analysis detected a greater number of high-risk interactions than the microsimulation model, suggesting that simulation tools may underestimate conflicts under heterogeneous traffic conditions. These findings highlight the importance of integrating empirical conflict observation with simulation-based analysis to support proactive traffic safety evaluation at unsignalized intersections in developing urban environments.