Human performance consequences of automated decision aids

作者
Reichenbach, Juliane
出处
期刊: 被引量:12
标识
DOI:10.14279/depositonce-4924
摘要

Introducing automation intends to yield performance benefits, increase system safety and efficiency, and decrease costs. However, the intended benefits are often offset by negative effects such as complacency and automation bias, loss of situation awareness, and skill degradation. Automating systems is not an all-or-none decision, and degree of automation (DOA) has been shown to influence both intended performance benefits as well as performance decrements. While performance benefits more from higher automation, results about the optimal degree of automation in terms of preventing negative effects are not so clear-cut. In a first study, we investigated effects of degree of automation on routine performance and failure performance using a simulated process control task. Results show that primary and secondary task performance as well as workload benefit from providing automation, and performance gains were higher for higher DOA. However, skill degradation was observed for the highest DOA when returning to manual performance. Regarding automation bias, there was no effect of DOA. The second study focused on the effects of system experience. Failure experience led to a strong decrease in trust, also reflected in more intense automation verification. Early failure experience reduced the risk of complacency and automation bias but did not prevent it completely. Moreover, we identified three possible causes of commission errors: a) incomplete automation verification, b) complete automation verification without attentive processing of contradictory information, analogous to a “looking-but-not-seeing” effect, c) discounting of contradictory system information. The third study looked into effects of operator functional state and DOA. Higher DOA could better protect performance after extended wakefulness, but return-to-manual performance suffered more. More information was sampled for automation verification during the night, and the risk of commission errors was lower compared to daytime performance.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
jinmai完成签到 ,获得积分10
刚刚
Hi完成签到,获得积分10
刚刚
薯条发布了新的文献求助10
刚刚
ww完成签到,获得积分10
刚刚
jr发布了新的文献求助10
1秒前
yh完成签到,获得积分10
1秒前
羲和完成签到,获得积分10
1秒前
丁驰完成签到,获得积分10
1秒前
rosyw完成签到,获得积分10
1秒前
yang完成签到,获得积分10
1秒前
ln177完成签到,获得积分20
1秒前
偷菜帅哥完成签到,获得积分10
1秒前
怡然的芯完成签到,获得积分10
1秒前
翻斗花园李元芳完成签到,获得积分10
2秒前
他说发布了新的文献求助10
2秒前
情怀的应助被WANGJD采纳,获得10
3秒前
3秒前
JLL完成签到 ,获得积分10
3秒前
木木完成签到,获得积分10
3秒前
zJx丶完成签到,获得积分10
3秒前
hhh发布了新的文献求助10
3秒前
EliasChan完成签到 ,获得积分10
3秒前
123完成签到,获得积分10
4秒前
芽芽配茄子完成签到,获得积分10
4秒前
123完成签到,获得积分10
4秒前
YPHCC发布了新的文献求助10
4秒前
4秒前
111完成签到,获得积分10
4秒前
山止川行完成签到 ,获得积分10
5秒前
隐形曼青的应助被锋zai采纳,获得10
5秒前
5秒前
记得吃早饭完成签到,获得积分10
5秒前
tfsn20完成签到,获得积分0
5秒前
Lucky晨完成签到,获得积分10
5秒前
绵马紫萁完成签到,获得积分10
5秒前
lisier完成签到,获得积分10
5秒前
6秒前
马嘉祺完成签到,获得积分10
6秒前
科研小白完成签到,获得积分10
7秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
Social Psychology in the Real World 600
八旗制度の研究 600
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8180275
求助须知:如何正确求助?哪些是违规求助? 9570366
关于积分的说明 21810725
捐赠科研通 7776380
什么是DOI,文献DOI怎么找? 3433277
关于科研通互助平台的介绍 2557483
邀请新用户注册赠送积分活动 2479042