Emotional evaluation of human arm motion models

Abstract

Emotional evaluation of human arm motion models is considered by the use of semantic differential test. The models are determined by changing the velocity pattern. The human arm motion model is recreated by CRT display and by an industry robot. In order to verify the influence of the smoothness of the velocity peak position to human motions, the robot hand motions whose velocity are based on the human results and the robot hand motions whose velocity are approximated by a triangle are conducted. The results show that the smoothness does not influence the evaluation, though that the influence from the velocity peak position and maximum velocity needs to be discussed. To examine the impression given to human by the velocity peak position, the motions whose velocity has different peak position are conducted about the triangular velocity pattern. The results show that human feel the most human-like to the motion whose velocity peak is in the first half of the duration and second most to that whose velocity peak is in the middle. However, human does not feel human-likeness to that in the second half of the duration. In order to examine the influence of maximum velocity, the motions whose velocity peak is in the first half of the duration are conducted under different maximum velocities. The results show that there exists proper maximum velocity which makes us feel human-likeness the most.<>