Traffic accidents annually result in millions of injuries and substantial economic losses worldwide. Historically, safety evaluations have relied on police-reported crash data to mitigate these incidents. However, this approach primarily reacts to accidents, making it less effective for proactive safety assessment. To enhance safety analysis, a more effective approach involves the use of conflict-based safety assessment with a range of surrogate measures of safety (SMoS). These measures allow for the study of critical traffic events that occur more frequently, aiding in the prevention of accidents. This study employs temporal proximity indicators, including Time to Collision (TTC), Post-Encroachment Time (PET), and Time Advantage (TAdv), to evaluate the safety of un-signalized intersections. Furthermore, investigate the critical speeds of vehicles involved in right-turning movements at these intersections to assess conflict severity, as deter-mined by PET values. Additionally, at these intersections, employ k-means, a centroid-based clustering method, to assess the severity levels of right-turning and straight-through traffic using TAdv and TTC, respectively. Our findings indicate that, within the selected study area, the SMoS at Krishna Arcade demonstrate a higher degree of risk compared to Sabban Choraha.