Department of Civil Engineering, McMaster University, 1280 Main Street West, Hamilton, Ontario L8S 4L7, Canada.
Accid Anal Prev. 2023 Nov;192:107229. doi: 10.1016/j.aap.2023.107229. Epub 2023 Aug 3.
Vision Zero stands out as one of the most promising systemic safety action plans. A crucial step to ensure the successful implementation of Vision Zero is to continuously assess the efficiency of the implemented treatments. Traditionally, this is achieved using before-and-after analyses or cross-sectional studies. However, the applicability of these approaches can be limited in assessing Vision Zero initiatives, which usually involve installing multiple treatments at a location, leading to a significant interdependency between treatments. This study proposes a dynamic R-vine copula-based time series model to evaluate the efficiency of treatments implemented as a part of Vision Zero. The proposed approach enables the accurate assessment of the treatments, understanding of their long-term impacts, and identifying the most effective combination of treatments at a location. The study also investigated the association between location characteristics and the performance of treatments. The proposed framework was applied to the City of Toronto at the macro-level (neighbourhood level) and focused on pedestrian-related treatments. Collision data and the implemented countermeasures were obtained from Toronto's Vision Zero Mapping Tool. The results show that the combination of speed limit reduction, leading pedestrian intervals (LPI), and community safety zones was the most frequent combination in terms of efficiency. Enforcement and speed limit reduction were the most effective combination in neighbourhoods with high school density, while LPI was effective in neighbourhoods with high densities of subway stations, and office density, especially when integrated with speed limit reduction and community zones. Driver feedback signs were effective in neighbourhoods with a high density of intersections, but only when combined with automated enforcement, community safety zones, and speed limit reduction. The results of the study would assist decision-makers in selecting the most effective treatment in a neighbourhood based on the neighbourhood characteristics and the countermeasures that are already installed.
零愿景是最有前途的系统安全行动计划之一。确保零愿景成功实施的关键步骤是不断评估已实施措施的效率。传统上，这是通过前后分析或横断面研究来实现的。然而，这些方法的适用性在评估零愿景计划时可能会受到限制，因为这些计划通常涉及在一个地点安装多个措施，导致措施之间存在显著的相互依存关系。本研究提出了一种基于动态 R-vine Copula 的时间序列模型，用于评估作为零愿景一部分实施的措施的效率。该方法可以准确评估措施的效果，了解其长期影响，并确定在一个地点最有效的措施组合。本研究还调查了地点特征与措施效果之间的关系。该框架应用于多伦多市的宏观层面（邻里层面），并专注于与行人相关的措施。碰撞数据和实施的对策是从多伦多的零愿景映射工具中获得的。结果表明，从效率的角度来看，限速降低、行人先行间隔（LPI）和社区安全区的组合是最常见的组合。在高中密度的社区中，执法和限速降低是最有效的组合，而在地铁站密度高、办公室密度高的社区中，LPI 是有效的，特别是与限速降低和社区区相结合时。驾驶员反馈标志在交叉口密度高的社区中有效，但只有与自动执法、社区安全区和限速降低相结合时才有效。该研究的结果将帮助决策者根据社区特征和已安装的对策，在社区中选择最有效的措施。
