Design of 22-DOF pneumatically actuated upper body for child android ‘Affetto’

Abstract

Child robots have been used in a lot of studies on human-robot social/physical interaction because they are suitable for safe and casual experiments. However, providing many compliant joints and lifelike exteriors to enhance their interaction potential is difficult because of the limited space available inside their bodies. In this paper, we propose an upper body structure that consists of slider crank and parallel mechanisms for linear actuators and serial mechanisms for rotary actuators. Such combinations of several joint mechanisms efficiently utilize the body space; in total, 22 degrees of freedoms were realized in an upper body space equivalent to that of an 80cm tall child. A pneumatic drive system was adopted in order to fully verify the behavioral performance of the body mechanism. The proposed redundant and compact upper body mechanism can be a platform for testing the effectiveness of future exteriors for the little child android 'Affetto', which was developed in order to integrate several key characteristics for achieving advanced human-robot interaction.

Investigation of Causal Relationship Between Touch Sensations of Robots and Personality Impressions by Path Analysis

“…A child-like android robot named Affetto [21] was set on a desk in front of a chair, as shown in Fig. 2.…”

Transfer Learning of Complex Motor Skills on the Humanoid Robot Affetto

“…In [12], parameterized motion primitives have been stored within a neural associative memory using non-linear embeddings of policy parameters. While learning in [8], [10], [12] was conducted offline on precollected data sets, [3], [4]. (b) Experimental setup for transfer learning, the robot has to generate drumming actions based on the perceived reflection of the drum in a mirror.…”

Appearance of a Robot Influences Causal Relationship between Touch Sensation and the Personality Impression

“…We analyzed the relationships between touch sensations and personality impressions of a robot [24], which was a child android robot called "Affetto" [7], by using a path and factor analyses method. They found several causal relationships between four components of touch sensations, i.e., Preference, Resilience, Smoothness, and Naturalness, and three components of personality impressions, i.e., Likability, Capability, and Vitality [24].…”