有条件自动驾驶中接管请求的告警内容需求分析——基于访谈和问卷的调查研究  

Research on Alert Content Requirements for Takeover Requests in Conditional Autonomous Driving:A Survey Based on Interviews and Questionnaires

作　　者：张澄澄 张警吁[2,3] 刘芝孜 陈星 马舒 ZHANG Cheng-cheng;ZHANG Jing-yu;LIU Zhi-zi;CHEN Xing;MA Shu(Department of Psychology,Zhejiang Sci-Tech University,Hangzhou 310018,China;CAS Key Laboratory of Behavioral Science,Institute of Psychology,Beijing 10086,China;Department of Psychology,University of Chinese Academy of Sciences,Beijing 100086,China;Chongqing Changan Automobile Co.Jianxin East Road,Jiangbei District,Chongqing 405808,China)

机构地区：[1]浙江理工大学理学院,杭州310018 [2]中科院心理研究所行为科学重点实验室,北京100086 [3]中国科学院大学心理学系,北京100086 [4]重庆长安汽车股份有限公司,重庆405808

出　　处：《人类工效学》2024年第4期35-41,共7页Chinese Journal of Ergonomics

基　　金：浙江省自然科学基金/探索项目Q(L020C090010);浙江理工大学科研基金(18062304-Y);浙江理工大学青年创新基金(22062311-Y);国家自然科学基金(T2192932);重庆市科技创新与应用发展重点专项(cstc2021jscx-dxwtBX0020)。

摘　　要：目的 调研在有条件的自动驾驶中,驾驶员在不同接管时间充足性、危险程度的接管场景中的告警信息内容需求和需求程度,形成告警内容框架。方法 采用桌面研究、专家、用户访谈的方式来搜集整理告警内容条目,随后通过线上问卷调研的方式,招募129名驾驶者(男61名;女68名),评估不同接管情景(时间充足性2*危险程度2)下对各告警内容的需求程度。结果 告警内容(F=5.138,P<.001,η_(p)^(2)=0.039)、时间充足性(F=13.042,P<0.001,η_(p)^(2)=0.092)、危险程度(F=14.679,P<0.001,η_(p)^(2)=0.103)主效应均显著。不论情境如何变化,对“不接管的后果”,“如果接管怎么应对”,“不接管车辆状态怎么变化”、“还有多长时间车辆状态将发生变化”的需求程度始终很高,其中对不接管后果的信息需求始终最高。在时间更紧迫、危险程度更高的情境下,对告警信息的需求程度均会显著增加。结论 从用户角度全面系统分析了对不同告警内容的需求程度,得出了“不接管的后果”,“如果接管怎么应对”,“不接管车辆状态怎么变化”、“还有多长时间车辆状态将发生变化”这四类在不同接管情境下需求程度始终较高的告警内容。Objective To explore the content requirements and the degree of these requirements for driver alerts in conditional autonomous driving scenarios with varying degrees of takeover time available and levels of danger,thereby establishing a framework for the content of these alerts.Methods Alerts content items were collected through desk research,consultations with experts,and interviews with users,followed by an online questionnaire survey gathering 129 participants(61 male,68 female)to assess the demand for different takeover scenarios,categorized by levels of time adequacy and urgent levels.Results Statistical analysis showed significant main effects for alert content(F=5.138,p<0.001,η_(p)^(2)=0.039),time adequacy(F=13.042,p<0.001,η_(p)^(2)=0.092),and urgent level(F=14.679,p<0.001,η_(p)^(2)=0.103).Regardless of scenario variations,the demand for information on the“consequences of non-takeover,”“response strategies upon takeover,”“changes in vehicle state in the absence of takeover,”and“remaining time before vehicle state change”consistently remains high,with the most significant demand being for knowledge regarding non-takeover consequences.There is an observed increase in the need for alert information under conditions of more pressing time constraints and higher urgent levels.Conclusion This study,from the user's perspective,comprehensively and systematically examines the demand levels for different types of alert content.It concludes that the content regarding the“consequences of non-takeover,”“what to do if taking over,”“changes in the vehicle state without takeover,”and“time remaining before the vehicle state changes”retains consistently high demand across various takeover scenarios.

关 键 词：交通工程 智能网联 人因失误 自动驾驶 接管请求 告警内容 接管情境 交通安全 事故预防 

分 类 号：U471.1[机械工程—车辆工程]