Scholarlink

<p>Roundabouts are increasingly used to improve operations and reduce severe crash outcomes; however, evidence remains limited on how operating regimes—captured by speed and exposure—relate to surrogate conflict measures. This study analyzes UAV videos from 14 urban roundabouts in three Iraqi cities, observed in three 20-min periods per site. Vehicle trajectories were extracted and processed to detect interactions using TTC ≤ 5 s, with a sensitivity check at TTC ≤ 1.5 s. Conflict frequency was modeled using negative binomial regression with log (exposure) as an offset. After exposure normalization, average speed was negatively associated with conflict rates (IRR per 1 km/h: 0.817 rear-end, 0.803 lane-change, 0.869 crossing; ≈ 13–20% lower rates). Illegal access points were associated with substantially higher crossing-conflict frequency (IRR = 2.55, ≈ + 155%). Conflict criticality was analyzed using site-level random-intercept mixed-effects models with a TTC-based continuous proxy, where higher speed was associated with larger mean TTC conditional on conflicts (≈ + 1.6% to + 2.5% per 1 km/h, depending on conflict type). These results describe conditional associations within observed operating regimes and do not imply that higher speeds are safer, since Safe System principles emphasize increasing injury risk with speed; surrogate conflict metrics should therefore be interpreted alongside speed management objectives.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Faster but safer? Exploring the impact of vehicle speed on surrogate safety measures at roundabouts

  • 期刊论文

  • 发表日期：

    2026年4月1日

查看全文
  • Moosa Al-Issawy,
  • Amin Mirza Boroujerdian

摘要

Roundabouts are increasingly used to improve operations and reduce severe crash outcomes; however, evidence remains limited on how operating regimes—captured by speed and exposure—relate to surrogate conflict measures. This study analyzes UAV videos from 14 urban roundabouts in three Iraqi cities, observed in three 20-min periods per site. Vehicle trajectories were extracted and processed to detect interactions using TTC ≤ 5 s, with a sensitivity check at TTC ≤ 1.5 s. Conflict frequency was modeled using negative binomial regression with log (exposure) as an offset. After exposure normalization, average speed was negatively associated with conflict rates (IRR per 1 km/h: 0.817 rear-end, 0.803 lane-change, 0.869 crossing; ≈ 13–20% lower rates). Illegal access points were associated with substantially higher crossing-conflict frequency (IRR = 2.55, ≈ + 155%). Conflict criticality was analyzed using site-level random-intercept mixed-effects models with a TTC-based continuous proxy, where higher speed was associated with larger mean TTC conditional on conflicts (≈ + 1.6% to + 2.5% per 1 km/h, depending on conflict type). These results describe conditional associations within observed operating regimes and do not imply that higher speeds are safer, since Safe System principles emphasize increasing injury risk with speed; surrogate conflict metrics should therefore be interpreted alongside speed management objectives.