A Study on the Factors That Affect the Occurrence of Crashes and Near-Crashes

Project Description The purpose of this effort is to support the use of various components of the SHRP 2 database by personnel from Honda R Co. (Honda). Data Request Scope The dataset encompassed all crashes, near-crashes, crash-relevant conflicts, non-subject conflicts, balanced-sample baselines, and additional baselines currently accessible on the InSight website (V3.0). This consisted of approximately 42,000 events. The duration of each event encompassed 20 seconds before precipitating event until 10 seconds after for crashes and near-crashes. Baselines were 21 seconds. Data Specification Specific variables delivered consisted of: o Front Video View o Rear Video View o Eye Glance Location (only for events for which this reduction has been completed and shared by the project’s funding source, currently consists of a few hundred events) o Hands Video View (only for events for which eye glance location is available) o Time Series (including all variables) o Post Crash Interview (including all variables) o Event Detail Table (including all variables) o Map Matching Data (this was not included for crashes, or for near-crashes and baselines that are close to the trip origin or destination) - In addition, VTTI provided data for the following tables: o Trip Summary Table (including all variables for ~42,000 events) o Vehicle Data  Vehicle Detail Table (including all variables) o Driver Data  Driver Demographic Questionnaire (including all variables)  Driving History Questionnaire (including all variables)  Driving Knowledge Survey (including all variables)  Visual and Cognitive Tests (including all variables)  Conner’s Continuous Performance Test (including all variables)  Clock Drawing Assessment (including all variables)  Physical Strength Tests (including all variables)  Barkley's ADHD Screening Test (including all variables)  Risk Perception Questionnaire (including all variables)  Risk Taking Questionnaire (including all variables)  Sensation Seeking Scale Survey (including all variables)  Driver Behavior Questionnaire (including all variables)  Medical Conditions Medications (including all variables)  Sleep Habits Questionnaire (including all variables)  Medical Conditions and Medications – Exit (including all variables)  Driver Exit Interview (including all variables) o VTTI also summarized and provided the total distance traveled by each participant项目说明 本项目旨在为本田R有限公司（Honda R Co.，以下简称本田）的工作人员提供SHRP 2数据库（SHRP 2 Database）各类组件的使用支持。 ## 数据请求范围 本数据集涵盖当前可在InSight网站（V3.0版本）获取的全部碰撞事件、险肇事故、碰撞相关冲突、非主体冲突、平衡样本基线及额外基线，总计约42000条事件。对于碰撞及险肇事故类事件，每条事件的时长覆盖触发事件前20秒至触发事件后10秒；基线样本的时长则为21秒。 ## 数据规范 本次交付的具体变量包括： - 前置视频画面 - 后置视频画面 - 视线注视位置（仅针对已完成该类数据处理并由项目资助方共享的事件，目前仅包含数百条事件） - 手部视频画面（仅针对已获取视线注视位置的事件） - 时间序列数据（包含全部变量） - 碰撞后访谈（包含全部变量） - 事件详情表（包含全部变量） - 地图匹配数据（不适用于碰撞事件，以及行程起点或终点附近的险肇事故与基线样本） 此外，弗吉尼亚理工大学运输研究所（Virginia Tech Transportation Institute，简称VTTI）还提供了以下数据表： - 行程汇总表（包含约42000条事件的全部变量） - 车辆数据 - 车辆详情表（包含全部变量） - 驾驶员数据 - 驾驶员人口统计学调查问卷（包含全部变量） - 驾驶经历调查问卷（包含全部变量） - 驾驶知识问卷（包含全部变量） - 视觉与认知测试（包含全部变量） - 康纳斯持续性能测试（Conner’s Continuous Performance Test，包含全部变量） - 时钟绘画评估（包含全部变量） - 身体机能测试（包含全部变量） - 巴克利多动症筛查测试（Barkley's ADHD Screening Test，包含全部变量） - 风险感知调查问卷（包含全部变量） - 风险承担调查问卷（包含全部变量） - 感觉寻求量表调查（包含全部变量） - 驾驶员行为问卷（包含全部变量） - 健康状况与用药情况（包含全部变量） - 睡眠习惯调查问卷（包含全部变量） - 健康状况与用药情况——离场调查（包含全部变量） - 驾驶员离场访谈（包含全部变量） VTTI还汇总并提供了每位参与者的总行驶里程数据。

搜集汇总

背景与挑战

背景概述

该数据集是SHRP2自然驾驶研究的一部分，旨在分析影响碰撞和近碰撞事件发生的因素。数据集包含约42,000个事件，涵盖碰撞、近碰撞、冲突和基线数据，并提供多模态数据，如视频、眼动追踪、时间序列和驾驶员问卷调查（包括人口统计、行为、医疗等），用于全面研究驾驶安全。