Naoki Saiwaki

A Two-Month Field Trial in an Elementary School for Long-Term Human–Robot Interaction

Interactive robots participating in our daily lives should have the fundamental ability to socially communicate with humans. In this paper, we propose a mechanism for two social communication abilities: forming long-term relationships and estimating friendly relationships among people. The mechanism for long-term relationships is based on three principles of behavior design. The robot we developed, Robovie, is able to interact with children in the same way as children do. Moreover, the mechanism is designed for long-term interaction along the following three design principles: (1) it calls children by name using radio frequency identification tags; (2) it adapts its interactive behaviors for each child based on a pseudo development mechanism; and (3) it confides its personal matters to the children who have interacted with the robot for an extended period of time. Regarding the estimation of friendly relationships, the robot assumes that people who spontaneously behave as a group together are friends. Then, by identifying each person in the interacting group around the robot, it estimates the relationships between them. We conducted a two-month field trial at an elementary school. An interactive humanoid robot, Robovie, was placed in a classroom at the school. The results of the field trial revealed that the robot successfully continued interacting with many children for two months, and seemed to have established friendly relationships with them. In addition, it demonstrated reasonable performance in identifying friendships among children. We believe that these results demonstrate the potential of current interactive robots to establish social relationships with humans in our daily lives.

A tactile synthesis method using multiple frequency vibrations for representing virtual touch

This paper presents a useful tactile display method that can control multiple tactile sensations such as roughness, pressure, and friction sensations using simple vibratory stimulations. Our concepts are based on two points: frequency range selection for making selective stimulation on different tactile receptors types and timing control of the stimulation in response to hand movements. The selective stimulations were realized by selecting reactive frequencies of vibratory stimulation based on temporal response characteristics of tactile receptors. For representing roughness sensation, vibrating frequencies were modulated in response to hand velocity considered as a temporal coding perception of FA I type receptor. Two reactive frequencies were also selected for representing pressure sensation and friction sensation corresponding to SA I and FA II type receptors respectively. A wearable tactile display using ICPF (ionic conducting polymer gel film) actuators verified our proposed methods. Finally, we conducted a total texture feeling display combined with our methods for roughness, pressure, and friction sensations in a parameterized manner. Comparison with real clothes showed that some combinations of multiple tactile sensations could express texture feels of the expected materials.

Friendly social robot that understands human's friendly relationships

This paper reports our novel approach to developing a social robot. The developed robot is able to identify relationships among humans from their physical behaviors as it establishes long-term relationships with humans. This interactive humanoid robot attracts humans to interact with it and, as a result, induces them to perform their group behaviors in front of it. The robot recognizes friendly relationships among humans by simultaneously identifying each person in the interacting group. We conducted a two-month experiment in an elementary school. As a result, the robot successfully continued friendly interaction with many children for the two months, and demonstrated reasonable performance in identifying friendships among children. We believe this ability to maintain friendly relationships with humans and to identify human relationships is essential to behaving socially.

Composition and Evaluation of the Humanlike Motions of an Android

Both appearance and motion are important considerations in designing robots for interaction with humans. In communication, humans sense intelligence in their conversation partner. We aim to develop an intelligent robot that communicates in the same intelligent and human-like way in daily life. We use the total Turing test to evaluate the human-likeness of the android. In this test, subjects are shown either the android or another human for a short time. Subjects are then asked whether it was a human being or not. The android used in this study is called ReplieeQ2, which has an appearance and motions that closely resemble those of a human being. We hypothesize that an android that is given more human-like motions will be identified as a human in the total Turing test more often. To create human-like motion we observed humans behavior when sitting. We focused on eye, eyelid, neck and chest motions and analyzed them in detail. The android motions were made from this analysis and used in the Total Turing Test. We prepared three conditions: a static android (SA), an android given natural motions (MA), and a human sitting naturally (SH). We expose a subject to one of these conditions for either one or two seconds. We then examine how often they correctly identify the condition as either human or robot. SH was identified as human most often, followed by MA and then SA, for both exposure times. MA is judged to be significantly more humanlike than SA, within the two-second experiment time. This suggests that natural motions contribute to the human-likeness of the android. We also investigate the contributions of individual natural motions to human-likeness and likeability by a pair comparison method. The results show that eyelid motion is the most effective factor for expressing human-likeness and likeability. By enhancing the androids expressive power, this work contributes to the development of humanoids that communicate naturally with human beings.

An approach to visualization of active position in brain by MEG

Reports a way to visualize the active position of the brain, and an improvement method to estimate the dipole of the MEG (magnetoencephalogram), which is the central part of the activation. In the past, at the Medical School of Osaka University, a method was used where an early-stage position was set manually and only its neighborhood was searched, but that method is less precise. On the other hand, our new method automatically estimates the dipole position and its moment for all positions of the brain. As a result, the measurement accuracy of the dipole position and moment was improved, and the position and moment derived from this calculation is medically adequate.

A method for supporting Web page design based on impression of Web page

This paper proposes an approach for Web page design support by focusing on the impression of Web page with respect to the selection of color and font. Previous approaches for Web page design support often focus on the automation of HTML coding and the pasting of graphics and icons. However, the selection of color and font also greatly affect the impression of Web page even with the same (information) content. The implemented system with the proposed support method tries to acquire the taste or preference of the user from the Web pages which are selected by him/her during net surfing and utilize it for Web page design with the preferred impression.

Design of interactive interfaces to detect conceptual difference

Conceptual difference among different people is a serious problem on a group knowledge acquisition system and a creative thinking support system in which different people are concerned. In this paper, we study conceptual difference among different people and describe a method of detecting it in the case that different symbols are used in the same meaning and/or the same symbols are used in different meanings. Conceptual difference is defined and the outline of our system is described. The detecting algorithm is designed, and then a prototype system and its evaluation is introduced.

Change detection of conditions of motorcycle for estimation of dynamics of ride-type interface

This paper describes a method for detecting the change in the conditions and behaviors of motorcycles from the motion data acquired by simplified measurement systems that estimate the dynamics of ride-type interfaces. In this method, motorcycle motion data, such as the acceleration and angular velocities, are determined using a microcontroller. Then, the changes in the conditions and behaviors of the vehicles, such as turning off the engine, beginning to run the engine, and braking, are detected by using the singular spectrum transformation of the motion data. Experimental results obtained with real motorcycles show the feasibility of the method.

Design of Wearable Interface Considering Touch Communications

Wearable computing has the image of the high-level information processing which uses the small-sized computer. However, a motion of the body is barred by the complexity of wiring or apparatus and the function of assistance of operation which is the function of original dress is disregarded in many cases. Moreover, it tends to be kept at a distance that externals look like the cyborg by a general person. On the other hand, the interface that uses the movement of the body recently is actively researched. Then, the prototype system of a good design and fit was developed for the college woman and evaluated in this research. The system has the touch interfaces that pay attention to touch communications and can do the music sharing and the voice conversation by two-way, and the switch can be done by touch.