Fig 1.
Circumplex model of emotion [47] utilized as the model of emotions for the robot.
Fig 2.
Robot and its vertical oscillation/transition.
(A) Robot utilized in the study (Locomotive-CommU), (B) Six different oscillations designed for vertical oscillation of the cylinder of the robot, (C) Three different movements implemented for vertical transition of the cylinder of the robot.
Fig 3.
Implemented vertical movements.
Examples of the vertical movements of the robot by combining the vertical oscillation and transition of the cylinder.
Fig 4.
Robot’s gesture for emotion expression.
Gestures elected for emotion expression of the robot for (A) short-term expressions, and (B) long-term emotions. The mentioned times in the time axis are the approximate ones and varies depends on the design of each gesture.
Table 1.
Result of the gesture contest.
Table 2.
Plan of the experiment for the number of subjects (short-term emotion).
Table 3.
Number of subjects for the experiment (short-term emotion).
Fig 5.
After watching the first video, the participants evaluated the expressivity of emotions of the robot in the video. Then, they watched the second video and evaluated the expressivity as well as the clarity of emotions of the robot in the video.
Table 4.
Plan of the experiment for the number of subjects (long-term emotion).
Table 5.
Number of subjects for the experiment (long-term emotion).
Fig 6.
Results for joy and anger in short-term expressions.
Score of the expressivity and clarity by different vertical transitions or oscillations of the robot in the case of the short-term representation of joy and anger: (A) expressivity of joy by vertical transition and (B) by vertical oscillation, (C) clarity of joy by vertical transition and (D) by vertical oscillation, (E) clarity of anger by vertical transition and (F) by vertical oscillation. Note that for the expressivity, the improvement of the score is shown as explained in the paper.
Fig 7.
Results for sadness and relief in short-term expressions.
Score of the expressivity and clarity by different vertical transitions or oscillations of the robot in the case of the short-term representation of sadness and relief: (A) expressivity of sadness by vertical transition and (B) by vertical oscillation, (C) clarity of sadness by vertical oscillation when the robot had a moving-up behavior and (D) by vertical transition when the robot had a fast oscillation, (E) clarity of relief by vertical transition. Note that for the expressivity, the improvement of the score is shown as explained in the paper.
Fig 8.
Results for different emotions in long-term expressions.
Score of the expressivity and clarity by different vertical oscillations of the robot in the case of the long-term representation of emotions: (A) expressivity of joy, (B) clarity of joy, (C) clarity of anger, (D) expressivity of sadness, (E) expressivity of relief, (F) clarity of relief. Note that for the expressivity, the improvement of the score is shown as explained in the paper.
Fig 9.
Effective movements for different emotions in short-term expressions.
Summary of the effective vertical movements on improvements of expressivity and clarity for short-term emotion expressions of the robot. The vertical transition and oscillation of the robot is denoted with Tr. and Osc. in the figure, respectively. Note that in the figure, if a movement such as A was effective under the condition B, it is noted as (B × A). Also, the notation no-osc. indicates the no-oscillation condition.
Fig 10.
Effective movements for different emotions in long-term expressions.
Summary of the effective vertical movements on improvements of expressivity and clarity for long-term emotion expressions of the robot. In the figure, the notation no-osc. indicates the no-oscillation condition.