Scholarlink

<p>Signalized midblock crosswalks (SMCs) are critical for pedestrian safety and mobility, yet their Level of Service (LOS) has not been well-defined in existing guidelines. Current LOS measures focus separately on either vehicles or pedestrians, overlooking integrated evaluation. This study proposes a novel LOS framework for SMCs based on vehicle and pedestrian delays, using an integrated approach of unsupervised learning (k-means clustering), supervised learning (support vector machines), and user perception analysis. Field data from eight SMCs in Hyderabad, India were analyzed to establish delay-based LOS thresholds for pedestrians (≤ 33 to  &gt; 70 s) and vehicles (≤ 15 to &gt; 32 s). An LOS warrant chart was developed to combine both criteria, ensuring that final LOS reflects the least favorable condition for any road user. The proposed framework demonstrated strong alignment with user perception (MCC &gt; 0.5) and high classification accuracy (up to 88.26%). This approach provides transportation professionals with a validated, comprehensive tool to assess and improve SMC operations, enhancing safety and efficiency under mixed traffic conditions.</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

Assessment of level of service at signalized midblock crosswalks under mixed traffic conditions

  • 期刊论文

  • 发表日期：

    2025年9月20日

查看全文
  • Sandeep Manthirikul,
  • Udit Jain,
  • Ankit Kathuria

摘要

Signalized midblock crosswalks (SMCs) are critical for pedestrian safety and mobility, yet their Level of Service (LOS) has not been well-defined in existing guidelines. Current LOS measures focus separately on either vehicles or pedestrians, overlooking integrated evaluation. This study proposes a novel LOS framework for SMCs based on vehicle and pedestrian delays, using an integrated approach of unsupervised learning (k-means clustering), supervised learning (support vector machines), and user perception analysis. Field data from eight SMCs in Hyderabad, India were analyzed to establish delay-based LOS thresholds for pedestrians (≤ 33 to  > 70 s) and vehicles (≤ 15 to > 32 s). An LOS warrant chart was developed to combine both criteria, ensuring that final LOS reflects the least favorable condition for any road user. The proposed framework demonstrated strong alignment with user perception (MCC > 0.5) and high classification accuracy (up to 88.26%). This approach provides transportation professionals with a validated, comprehensive tool to assess and improve SMC operations, enhancing safety and efficiency under mixed traffic conditions.