Investigating the effects of silent automation failure and scenario criticality on automated vehicle’s takeover performance (03-036)

Project Description: Research question: Investigating the effects of silent automation failure and scenario criticality on automated vehicle’s takeover performance. Data collection: The study was conducted in July-August 2019 in the Texas A Transportation Institute (TTI) driving simulation lab. The lab consists of a Realtime Technologies Inc. (RTI) quarter-cab driving simulator with three screens that provide 165° horizontal and 35° vertical fields of view, a speaker system to provide ambient roadway noise, and a physiological and eye-tracking data collection suite. Data scope: The simulator collects kinematic time series data including steering wheel position, accelerator and brake pedal positions, velocity, time to lane crossing, time headway to an upstream object, and lane position at a 60 Hz sampling rate. The data was collected from 64 participants and included 4 experimental drives per participant. The driving simulator experiment used a 2×2×2 factorial design including a between-subjects factor of takeover request presence (silent vs. alerted) and within-subjects factors of scenario criticality (critical vs. non-critical) and takeover scenario (obstacle reveal and unexpected braking). Data Specification: Please see attached.

项目概述：
研究问题：探究静默自动化故障与场景临界性对自动驾驶车辆接管性能的影响。
数据收集：本研究于2019年7月至8月在德州A交通研究所（Texas A Transportation Institute, TTI）驾驶模拟实验室开展。该实验室配备实时技术公司（Realtime Technologies Inc., RTI）的四分之一舱式驾驶模拟器，其搭载三块屏幕，可提供165°水平视场角与35°垂直视场角，配套音响系统以还原道路环境噪音，同时集成生理与眼动数据采集套件。
数据范围：模拟器以60Hz的采样率采集运动学时序数据，具体包括方向盘位置、加速与制动踏板开度、行驶速度、车道穿越时间、前车时距以及车道位置。本次研究共招募64名受试者，每名受试者完成4次模拟驾驶实验。本次驾驶模拟实验采用2×2×2析因设计，其中被试间因子为接管请求触发方式（静默式 vs 告警式），被试内因子分别为场景临界性（临界场景 vs 非临界场景）与接管场景（障碍物突现与意外制动）。
数据规格说明：详见附件。