System Transparency in Shared Autonomy: A Mini Review

被引:75
作者
Alonso, Victoria [1 ]
de la Puente, Paloma [2 ,3 ]
机构
[1] Univ Politecn Madrid, ETSI Aeronaut & Espacio, Madrid, Spain
[2] Univ Politecn Madrid, ETSI Ind, Madrid, Spain
[3] Univ Politecn Madrid, CSIC, Ctr Automat & Robot, Madrid, Spain
关键词
transparency; shared autonomy; human-robot interaction; communication; observability; predictability; interface; user-centered design; HUMAN-ROBOT INTERACTION; BILATERAL TELEOPERATION; SITUATION AWARENESS; TEACHING ROBOTS; AUTOMATION; IMPROVE; COMMUNICATION; PERFORMANCE; STABILITY; MOTION;
D O I
10.3389/fnbot.2018.00083
中图分类号
TP18 [人工智能理论];
学科分类号
140502 [人工智能];
摘要
What does transparency mean in a shared autonomy framework? Different ways of understanding system transparency in human-robot interaction can be found in the state of the art. In one of the most common interpretations of the term, transparency is the observability and predictability of the system behavior, the understanding of what the system is doing, why, and what it will do next. Since the main methods to improve this kind of transparency are based on interface design and training, transparency is usually considered a property of such interfaces, while natural language explanations are a popular way to achieve transparent interfaces. Mechanical transparency is the robot capacity to follow human movements without human-perceptible resistive forces. Transparency improves system performance, helping reduce human errors, and builds trust in the system. One of the principles of user-centered design is to keep the user aware of the state of the system: a transparent design is a user-centered design. This article presents a review of the definitions and methods to improve transparency for applications with different interaction requirements and autonomy degrees, in order to clarify the role of transparency in shared autonomy, as well as to identify research gaps and potential future developments.
引用
收藏
页数:11
相关论文
共 169 条
[1]
Hybrid FES–robotic gait rehabilitation technologies: a review on mechanical design, actuation, and control strategies [J].
Anaya F. ;
Thangavel P. ;
Yu H. .
International Journal of Intelligent Robotics and Applications, 2018, 2 (01) :1-28
[2]
[Anonymous], 2017, 26 IEEE INT S ROB HU
[3]
[Anonymous], 2016, TRANSPARENT COMPUTER
[4]
[Anonymous], 2006, Def
[5]
[Anonymous], 2016, P IJCAI WORKSH ETH A
[6]
[Anonymous], 2018 IEEE RSJ INT C
[7]
[Anonymous], DELEGATION TRANSPARE
[8]
[Anonymous], THESIS
[9]
[Anonymous], 2006, ROMAN 2006, DOI DOI 10.1109/ROMAN.2006.314398
[10]
[Anonymous], 2016, Explainable Artificial Intelligence (XAI)