Impact of Roadway Lighting on Nighttime Crash Performance and Driver Behavior

Project Description The goal of this SHRP 2 Project is to explore the feasibility of using SHRP 2 NDS, RID, and CIBSS lighting data, for meaningful evaluation of the impact of lighting on night-time crashes and near-miss events. Findings of the analysis will have the potential of leading to lighting-related countermeasures that specifically address lighting characteristics (for example, the level of lighting (vertical and horizontal illuminance), roadway luminance, spot lighting, and lighting uniformity), lighting and policy/standard changes aimed at improving safety performance, reducing energy consumption and optimizing asset investments.. It is expected that the research will look at: The effects of roadway lighting characteristics on driver behaviors that affect safety, such as workload; The effects of roadway lighting levels on safety for different roadway geometries and traffic-control settings; and The recommended lighting levels needed to support safety performance in different roadway geometries and traffic-control settings. Data Request Scope The research team will request NDS data for the following roadways (may require additional roadways to ensure the identification of interesting lighting conditions) in the Seattle-TacomaBellevue metropolitan area: IH-5 and IH-405. The facilities pass through several interchanges and has a great deal of diversity in lighting levels (including segments with and without lighting). The facilities include a wide range of different lane configurations and represents a variety of traffic control settings. The sections will not be longer than 30 miles for each corridor, SR-522 from US-2 at Monroe and IH-5 at Seattle. This is a 24-mile surface arterial passes through several grade-separated or at-grade intersections. The roadway connects Monroe and Seattle and includes two-lane, four-lane, and six-lane cross sections with both controlled-access and non-controlled-access segments. The roadway also includes segments with and without street lighting. Up to 20 intersections in the Seattle area will be selected. Data Specification Time Series data points for a total of up to 1000 trips for each of the three corridors mentioned above including the following variables (may request additional variables as needed): Speed, GPS; Speed, Vehicle Network; Acceleration, x-axis; Acceleration, yaxis; Yaw Rate, z-axis; Turn Signal; Acceleration, z-axis; Cruise Control; Day; Dilution of Precision, Position; Head Confidence; Head Position X; Head Position X Baseline; Head Position Y; Head Position Y Baseline; Head Position Z; Head Position Z Baseline; Head Rotation X; Head Rotation X Baseline; Head Rotation Y; Head Rotation Y Baseline; Head Rotation Z; Head Rotation Z Baseline; Illuminance, Ambient; Lane Marking, Distance, Left; Lane Marking, Distance, Right; Lane Marking, Probability, Right; Lane Marking, Type, Left; Lane Marking, Type, Right; Lane Markings, Probability, Left; Lane Position Offset; Lane Width; Latitude; Location; Longitude; Month; Pitch Rate, y-axis; Pitch Rate, y-axis fast; Radar, Range Rate Forward X for Tracks 0 to 7; Radar, Range Rate Forward Y for Tracks 0 to 7; Radar, Range, Forward X for Tracks 0 to 7 ; Radar, Range, Forward Y for Tracks 0 to 7; Radar, Target Identification for Tracks 0 to 7; Roll Rate, x-axis; Roll Rate, x-axis fast; Subject_ID; Time; Timestamp; vehicle_id; Yaw Rate, z-axis fast; and Year. Intersections. The research team will provide VTTI SHRP2 data team with a list of up to 50 intersections represented in the Washington state lighting database. For these intersections, VTTI will provide a summary table of characteristics with the following variables: Number of trips through intersection Number of drivers who drove through the intersection Age bins and gender distributions for those drivers Time of day distributions Number of crash and near crashes if any Event Detail data. The research team will request the event data table for night crashes and near crashes for the State of Washington. Depending on the data and time availability the research team will select 50 to 100 events that will be reviewed in greater detail in the secure data enclave. For this limited number of events baseline events will also be required at this point. Potential variables to review on include Video Dashboard and Steering Wheel View; Video Frame; Video, Driver and Left Side View; Video, Forward Roadway; Video, Occupancy Snapshot; Video, Rear View. Driver Demographic Questionnaire and Driver Behavior Questionnaire data for the drivers that are involved in the extracted Time Series and Event Detail data. Video data for selected trip epochs. The research team will look at video data for up to 250 special epochs of the selected final trips. Potential variables that the research team will look at are the same described above for the events.

项目描述
本SHRP 2项目的核心目标是探索利用SHRP 2的全国驾驶数据集（National Driving Dataset, NDS）、道路照度数据（Road Illuminance Data, RID）以及CIBSS照明数据，开展照明对夜间碰撞与险肇事件影响的系统性评估的可行性。本分析所得结论有望催生针对性的照明相关干预措施，精准覆盖各类照明特性（例如照明水平——包括垂直与水平照度、道路亮度、重点照明以及照明均匀度），以及旨在提升安全绩效、降低能耗并优化资产投入的照明与政策/标准调整方案。本研究预计将围绕以下方向展开：
1. 道路照明特性对影响安全的驾驶员行为（如认知负荷）的影响；
2. 不同道路几何形态与交通管控场景下，道路照明水平对安全的影响；
3. 不同道路几何形态与交通管控场景下，支撑安全绩效所需的推荐照明水平。

数据请求范围
研究团队将申请西雅图-塔科马-贝尔维尤都会区内以下道路的NDS数据（可能需新增道路以确保可识别出典型照明场景）：I-5州际公路与I-405州际公路。该两条公路途经多处互通式立交，照明水平差异显著（涵盖有照明与无照明路段），且包含多种车道配置类型，覆盖多样的交通管控场景。每条走廊的路段长度均不超过30英里；此外还有门罗市2号美国国道段与西雅图I-5段相连的522号州道：该公路为全长24英里的地面主干道，途经多处立体交叉或平面交叉口，连接门罗市与西雅图，包含双向两车道、四车道及六车道的横断面形式，同时涵盖受控通行与非受控通行路段，且包含有照明与无照明的街道照明区段。西雅图地区将选取至多20个交叉口。

数据规格
上述三条走廊的总计至多1000次行程将被采集时序数据点，包含以下变量（按需可申请新增变量）：GPS速度、车辆网络速度、x轴加速度、y轴加速度、z轴横摆角速度、转向灯状态、z轴加速度、巡航控制状态、日期、位置精度稀释度、头部检测置信度、头部X轴位置、头部X轴基准位置、头部Y轴位置、头部Y轴基准位置、头部Z轴位置、头部Z轴基准位置、头部X轴旋转角度、头部X轴基准旋转角度、头部Y轴旋转角度、头部Y轴基准旋转角度、头部Z轴旋转角度、头部Z轴基准旋转角度、环境照度、左侧车道标线距离、右侧车道标线距离、右侧车道标线存在概率、左侧车道标线类型、右侧车道标线类型、左侧车道标线存在概率、车道位置偏移量、车道宽度、纬度、定位信息、经度、月份、y轴俯仰角速度、快速y轴俯仰角速度、0至7号跟踪目标的前向X轴雷达距离变化率、0至7号跟踪目标的前向Y轴雷达距离变化率、0至7号跟踪目标的前向X轴雷达探测距离、0至7号跟踪目标的前向Y轴雷达探测距离、0至7号跟踪目标的雷达目标识别信息、x轴侧滚角速度、快速x轴侧滚角速度、受试者编号、时间、时间戳、车辆编号、快速z轴横摆角速度以及年份。

交叉口相关数据
研究团队将向VTTI SHRP2数据团队提供华盛顿州照明数据库中涵盖的至多50个交叉口列表。针对这些交叉口，VTTI将提供包含以下变量的特征汇总表：交叉口通行总次数、途经该交叉口的驾驶员人数、该类驾驶员的年龄分组与性别分布、时段分布情况、碰撞与险肇事件数量（若有）以及事件详情数据。
研究团队将申请华盛顿州夜间碰撞与险肇事件的事件数据表。根据数据与时间可用性，研究团队将选取50至100起事件，在保密数据专区内开展深度复盘。针对该有限数量的事件，同时还需获取基准事件数据。待复盘的潜在变量包括：仪表盘与方向盘视角视频、视频帧、驾驶员与左侧视角视频、前方道路视角视频、占用快照视频、后方视角视频。
针对提取的时序数据与事件详情数据所涉及的驾驶员，收集其驾驶员人口统计问卷与驾驶员行为问卷数据。

选取行程时段的视频数据
研究团队将对选定最终行程的至多250个特殊时段开展视频数据分析，待分析的潜在变量与前述事件分析所用变量一致。

搜集汇总

背景与挑战

背景概述

该数据集来自SHRP 2项目，旨在研究道路照明特性（如照明水平和均匀性）对夜间事故和驾驶员行为的影响，以推动安全改进和能源优化。数据集包括西雅图-塔科马-贝尔维尤都市区特定道路的时间序列数据（如速度、光照度）、交叉路口特征、事件详情以及驾驶员问卷，覆盖多种道路几何和交通控制设置，用于分析照明与安全性能的关联。