Advancing knowledge of the safety implications of transportation engineering decisions with surrogate safety data and innovative statistical techniques

近年来，交通工程师研究和评估道路和交通安全的方法有了显著的进步。工程师现在可以采用系统和严格的流程来分析道路现场的安全性，以确定减少碰撞的机会，并实施潜在的安全对策。这一过程主要依赖于一套经过多年研究开发的统计技术和模型。反过来，这些模型依赖于碰撞观察作为量化安全水平的主要数据源。然而，在安全模型中使用碰撞数据有一些局限性，主要与质量和可用性问题有关。此外，使用崩溃观察是被动的：在采取任何行动之前，必须发生大量的崩溃。这些限制也阻碍了该领域方法论进步所提供的新机会的应用。

The way transportation engineers study and evaluate road and traffic safety has remarkably advanced in recent years. Engineers can now employ a systematic and rigorous process to analyze safety at road sites that identifies opportunities to reduce crashes and implement potential safety countermeasures. This process mainly relies on a set of statistical techniques and models developed over several years of research. In turn, these models rely on crash observations as the main data source to quantify safety levels. However, the use of crash data in safety models has several limitations, mainly related to quality and availability issues. Moreover, the use of crash observations is reactive: a substantial number of crashes must occur before any action is taken. These limitations also inhibit application of new opportunities offered by methodological advancements in the field.