Delays of response in Human-Robot Interaction
|has title::Delays of response in Human-Robot Interaction|
|Master:||project within::Artificial Intelligence (Cognitive Science)|
|Student name:||student name::Jorge Gallego Pérez|
|Company:||has company::University of Plymouth|
Abstract KIM 2
Embedded robots (i.e. robots that are influenced and can influence their physical environment to a high degree) have served several purposes in society since their first appearance. Clear examples are industry automation or space exploration. Far from confined to science fiction movies and speculation, the field of social robotics represents a reality. Human-Robot Interaction (HRI) studies can emphasize the role of the robot (robot centered HRI), its socio-cognitive abilities (robot-cognition centered HRI) or, as it is the scope of the present study, the role that the human plays in the interaction (human centered HRI).
A NAO robot (Aldebaran Robotics) was shown to youngsters at a school science fair in Exeter, United Kingdom. A natural, not constrained interaction took place between NAO and the children that spontaneously decided to play with it. Among other behaviors, NAO reacted to touch on its head by displaying funny sounds. This reaction, however, could occur immediately after touch or after a certain delay. The present study focuses on how these delays alter the quality of the interaction.
Objective measures were taken (data from the robot´s sensors and video recording), as well as subjective (rating about the quality of those interactions, assessed by external observers who watched the corresponding videos). Among others, the following findings were obtained:
- Not surprisingly, the overall tendency is that the longer the delay is, the worst the quality of the interaction becomes.
- However, interaction seems to become suddenly very poor at delays of 0.5 s and to recover again at 1 s.
More results and their corresponding argumentation will be discussed.