Keywords

1 Introduction

The processes of how human emotions arise and functions of emotions have been studied since ancient times. Recently, in the field of information engineering, there have been many studies to measure and control emotions. Among the basic human emotions as classified by Ekman, we focus on fear or anxiety [1, 2].

Fear or anxiety are closely associated with anxiety disorder. Approximately 40% of cases of depression is preceded by anxiety disorder, and anxiety disorder is closely associated with increasing societal cost of mental disorders [3]. The Ministry of Health, Labour and Welfare [4] in the 2014 patient survey estimated the number of patients with depression to be 112,000, and the number of patients with neurotic disorder to be more than 59,000. These numbers have been increasing since 1996. Cabinet Office, Government of Japan [5] has reported that based on the Japanese citizen poll conducted in June 2014, approximately 67% of Japanese people have worry or anxiety in their daily lives.

When human is contact with a person physically, his/her fear or anxiety will be reduced as Gergen et al. [6] shows. In this research [6], people were divided into several groups. Each group contained approximately eight persons. Some groups were left in the bright room and other groups were left in dark room. In the dark room, the people lost spatial sense of direction and visual communication with others became impossible. Although the subjects did not know each other, approximately 90% of them intentionally sought physical contact with others in the dark room. Furthermore, approximately 50% of the subjects hugged each other. In the bright room, there was no physical contact or hug. In response to this experimental result, Yamaguchi [7] suggested that the physical contact can reduce humans’ anxiety. Further-more, the experimental results that showed the human’s fear or anxiety will be reduced when he/she is contacted with other person physically. The person is not need to be a specific people, such as close friends, a romantic partner, or family. Even physical contact with an unrelated person can reduce fear or anxiety.

Although physical contact with an unrelated person is successful in reducing fear or anxiety in a dark environment, that person is usually someone in close relationship such as friends, romantic partners, and family. Hall [8] categorized interpersonal distance into four categories—close, personal, social, and public distances—and describes how physical distance between people communicating with each other is proportional to psychological distance. This suggests that physical contact with an unrelated person is hard to establish and in usual cases the physical contact is established with the specific person such as friends, romantic partners, and family. Since unmarried people are increasing [9] and single-person households are increasing [10] in Japan recently, it is becoming difficult to realize physical contact in daily life.

In associated with the above problem, there are attempts to support communication between specific people in some fields including tele-existence [11]. However there are few attempts to support communication between unrelated people and to validate its effect, although fear or anxiety will be reduced by communication with unrelated person.

Therefore, we attempt to reduce fear or anxiety by supporting physical contact with unrelated person instead of physical contact with specific person. We developed “Breath Chair,” a chair system that aims to reduce fear or anxiety [12] (Fig. 1). It simulates the movement of a person’s thorax when he/she is breathing. When the user sit in this chair system, the user feels the sense as if the user is physically contact with him/her. In this paper, we refer to the movement of the thorax during breathing simply as “breath.” The objective of this paper is to perform an experiment using Breath Chair and show that fear or anxiety are reduced by physical contact via simulated breath. In addition, we aimed to show that fear or anxiety are reduced by physical contact with one, even if one is an unrelated person to the subject.

Fig. 1.
figure 1

Using Breath Chair

Full size image

2 Related Work

In this Section, we will discuss related works on transmission of physiological information and to affect emotions, followed by tele-existence and communication.

2.1 Transmission of Physiological Information and Effect on Emotions

The relationship between changes in emotion and physiological responses has been discussed for a long time [13]. For change in emotions and physiological responses, Nishimura et al. [14] aimed to artificially manipulate fondness for people by externally reinforcing or suppressing changes in physiological responses of liking people and developed a device that provides tactile presentation of one’s own heartbeat.

“Lovable Couch” developed by Iwamoto et al. [15] visualizes heartbeat as an information to judge the level of romantic interest between a male and a female who met for the first time. By visualizing the heartbeat of the communication partner, it provides a factor for judging the level of interest expressed by the communication partner. In this manner, it experimentally derives a positive correlation between the actual affection and heartbeat information. Based on this result, it supports judgement of affection expressed by the other sex and communication.

As such, there are attempts to support good communication by visualizing physiological information and changes at a specific emotion. However, while there have been studies to understand the relationship between a specific emotion and changes in physiological response, not many studies have examined the effects of simulating constant physiological information under normal and rested states. For example, the vibration from heart beat and the movement of the thorax during breathing. In the estimation and evaluation of emotions and associated physiological response, not only are happiness and fear, or positive and negative emotions, but also neutral state and emotion are assumed [16, 17]. Yet, in visualizing emotions and physiological information, neutral state and emotion and their physiological information have not been studied much. Thus, we attempt to examine the effects of simulating constant physiological information under normal and rested states as well as under neutral state and emotions.

2.2 Tele-Existence in Communication

Tanaka et al. [11] added remote contact using a robot hand for handshaking to a conventional video conference system with audiovisual information in order to strengthen social tele-presence.

Sakamoto et al. [18] performed an experiment in which the experimenter and subjects have a conversation. This experiment evaluated the impression of the sense of presence under three different conditions: with a remotely operated robot modeled after and closely resembling a real person, with a video conference system, or with a speaker like a telephone.

Yamaoka et al. [19] used an autonomous robot and created two conditions: one where subjects were told that robot movements were based on a program and another where subjects were told that robot movements are controlled by an operator. In this manner, they examined differences in impression of the robot. The result showed that regardless of the condition, 2/3 of subjects felt that they were interacting with the robot itself.

As shown by these studies, the sense of presence with video conference system and robots are being examined. When a communication is made through a remotely operated robot, does the user recognize the communication as interaction with the robot or with the operator? However, in many cases, most subjects assume that they are communicating with a specific person under a specified situation. When using a video conference system, visual information clearly defines the communication partner. Even when using a robot, most subjects feel that they are interacting with the robot itself, regardless of the condition; thus, it is assumed that the robot in front of the subject is the communication partner.

Although fear or anxiety can be reduced via physical contact with an unrelated person, there have been few efforts and studies to support and validate the value of communication with an unrelated person. Thus, in this study, the proposed system will not provide physical characteristics or information that will allow subjects to identify an actual person and will define such object without identity as an unrelated person. We will then examine if fear or anxiety can be reduced by having simulated physical contact with such unrelated person.

3 Breathing Movement Simulation System

We developed a chair system (called “Breath Chair”) which simulates the movement of a person’s thorax when he/she is breathing. When the user sit in this chair system, the user feels the sense as if the user is physically contact with him/her [12]. In this study, we assume simulated physical contact with an unrelated person. If we prepare the robot-type medium with only torso (without head or limbs), the subjects do not identify specified communication partner and are not influenced by the shape of the robot-type medium [20]. Also considering the lifestyle of Japanese people [21], we designed our system as chair-style system which has only chest and which does not have head, limb, and so on.

The schematics of this system is shown in Fig. 2. This system has sealed polyurethane sponge in the back of the chair. This polyurethane sponge is compressed and expanded with a vacuum pump, and these volume changes simulate breathing, providing a sense of physical contact to reduce fear or anxiety.

Fig. 2.
figure 2

System configuration

Full size image

The objective of this paper was to show that fear or anxiety are reduced by physical contact through simulated breathing, even if the physical contact is made with a person unrelated to the subject. Breathing in this system was visualized through controlled volume change of polyurethane sponge using compression and expansion based on average adult respiratory rate to avoid presenting characteristics such as sex and age of a person by basing on the breathing of an actual person.

Based on the reports that chest expansion, or the difference in chest circumference between maximal inhalation and exhalation, is approximately 3.3 cm [22], volume change of polyurethane sponge between compression and decompression was controlled to approximately 3.3 cm in perimeter while the seat was vacant. Therefore, changes in perimeter occur when a user is seated in the present system due to changes in pressure on the back of the chair depending on the user’s weight and the way the user sits. Since, the rate of breathing in this system and ratio of rhythm associated with inhalation and exhalation were designed based on the average adult’s respiratory rate [23], the rhythm polyurethane sponge’s movement was regulated 12 times in a minute. Since the ratio of inspiration and expiration in adults is 1:3, we set the rhythm of inhalation and discharge of air in this system at 1:3.

4 Experiment: Effect on User’s Fear and Anxiety

We conducted experiments using our system “Breath Chair” which aims to reduce fear or anxiety. The objective of these experiments was to show that physical contact using simulated breathing can reduce fear or anxiety. We also aimed to show that fear or anxiety are reduced even if simulated physical contact is caused as breathing movement of an unrelated person. When conducting such experiments, the recognition for the system’s movement by the subject must be distinguished. We must distinguish whether the subject recognizes the system’s movement as the movement of the thorax during human breathing or the subject recognizes it as a simple physical movement. Thus, we conducted the following two identical experiments. Between two experiments, our system Breath Chair gave same movements to subjects, but information provided to the subjects differed as follows. In Experiment 1, we informed each subject that Breath Chair’s movements were only physical movement and consequent external stimulation. In Experiment 2, we informed each subject that Breath Chair’s movements were breathing movements which were simulated by measuring someone’s breath who was located at a distance through real-time remote sensing. Although the true objective of our experiments was to reduce fear or anxiety using our system “Breath Chair,” in order to remove bias in the subjectivity and recognition of every subject, we informed each subject that “The aim of these experiments is to examine the effects of our system Breath Chair regardless whether the effect is positive or negative” at the beginning of experiments. After these experiments were completed, we explained the true objective of our study to each subject.

In this chapter, we first noted the common contents between Experiment 1 and Experiment 2. Then, we discuss Experiment 1 in Sect. 4.6 and Experiment 2 in Sect. 4.7.

4.1 The Subjects and the Experimental Environment

In our experiments, we used a questionnaire, State-Trait Anxiety Inventory (STAI), that evaluates anxiety by dividing anxiety into state and trait anxieties. Since, Spielberger et al. [24], who developed the STAI conducted his experiment on university students, we also selected university students and graduate students as subjects of our experiment. We informed the objective and methods of our experiments to these subjects. Also each subject was informed that there is no disadvantage if they reject our experiments, that they are free to discontinue our experiments at any moment, that obtained data are statistically processed, that there is no identification of individuals in analysis, and that privacy is protected, by our verbal explanation and a written format. In this manner, informed consent was obtained. This study was approved by the Ethical Review Board for the use of human subjects of Kanagawa Institute of Technology.

The experiment used an air-conditioned room into which other people did not enter or leave. Based on the experiment conducted by Honda et al. [25], the lighting in the room was set to approximately 30 lx. To eliminate factors affecting the sympathetic nervous system as much as possible, each subject was instructed to refrain from alcohol the day before, exercise on the day in question, and was not given any food except for water 2 h before the start of the experiment. A visual stimulation was presented by placing a 23-inch display approximately 1 m in front of the test subject. During the experiment in the test room, the examiner was absent, and the test subject was left alone.

4.2 Evaluation Indicators

In these experiments, we used both psychological indicator and physiological indicator. The psychological indicator STAI (state anxiety scale) was used as an anxiety indicator. The STAI includes scales for both the state anxieties and the trait anxieties. As STAI expresses anxiety, the higher the value, the greater the anxiety. The distribution of STAI value occurs within the range of 20 and 80 points. State anxiety is a transient state reaction to the phenomenon invoking the anxiety and it measures “what I am actually feeling now.” On the contrary, trait anxiety measures long-term result and indicates “how do I feel normally in general.” Based on the STAI created by Spielberger et al. [24], Hidano et al. [26] developed new items to consider Japanese culture and checked its reliability. We used Hidano’s STAI in these experiments.

As a physiological indicator, the fingertip surface skin temperature was used to indicate fear or anxiety. The fingertip surface skin temperature was calculated as the average value for every 30 s, using a thermistor thermometer (NXFT15XH103FA2B) attached to the fingertip ventral section of the index finger of the left hand. When the sympathetic nervous system is excited, that excitation constricts the peripheral blood vessels and reduces blood flow, then it decreases the skin surface temperature in the peripheral area. Such decrease in the skin surface temperature was also reported by Kumamoto et al. [27] in their study, where the perception level for pain, etc., and psychological anxiety increased.

4.3 Stimulus Invoking Emotion Such as Fear or Anxiety

In order to invoke emotion such as fear or anxiety to each subject, we gave load stimulus to the subject. Although visual stimulus is not invasive to subject, it invokes comparatively strong emotion to the subject in the test room. Many researchers [28, 29] studied visual stimuli which invoke specific emotions to subject. Furthermore, Honda et al. [HMY02] reported that if the subject is given visual stimulus, the fingertip skin surface temperature of the subject changes significantly. Based on these findings, we gave visual stimulus to each subject in order to invoke fear or anxiety to each subject in our experiments.

As visual stimuli in our experiment, we used the same images which Alexandre et al. [30] used in their experiment, in order to invoke emotion such as fear or anxiety to each subject. We used the same videos which were exactly the same movies used by Alexandre et al. with the same start point and stop point, but we used the Japanese versions of the movies. In their experiment, the researchers used movies of virtually the same length, 210 s, which none of the test subjects in the experiment had seen up to that point. The two movies used were “Misery” (copyright; Castle Rock Entertainment, 1990) and “Scream 2” (copyright; Miramax Film Corp., 1997).

4.4 Experimental Protocol

Figure 3 shows the procedures of Experiment 1 and Experiment 2. In these experiments, each subject sat in our chair system Breath Chair under the experimental environment and relaxed for 3 min with his (her) eyes closed. Then, we measured the subject’s emotion by the questionnaires of STAI state anxiety scale and STAI trait anxiety scale. Next, as intervention 1, each subject viewed a video. Subsequently, we measured the subject’s emotion by the questionnaires of STAI state anxiety scale. Then the subject relaxed for a total of 6 min. Intervention 2 involved the same operations as intervention 1. After these experiments were completed, we asked each subject about impressions and opinions to the experiment. In these experiments, if our system Breath Chair worked during intervention 1, it did not work during intervention 2. If our system Breath Chair worked during intervention 2, it did not work during intervention 1. We assigned these combinations of Breath Chair On and Breath Chair Off randomly, in order to counterbalance the order effect and interaction. Also we assigned the combinations of two kinds of video randomly, in order to counterbalance the order effect and interaction.

Fig. 3.
figure 3

Experimental protocol

Full size image

4.5 Analytical Methods

To analyze the result of the questionnaires of STAI state anxiety scale, we performed one factor analysis of variance. For the fingertip surface skin temperature, we calculated the mean for each 30-s interval based on the experiment by Honda et al. [25]. We calculated each baseline as the mean of values which were measured when each subject relaxed before viewing each video in interventions 1 and 2. Then we calculated the amount of change by subtracting the baseline from value which was measured when the subject viewed each video in interventions 1 and 2. For multiple comparison of each variance analysis, we used Tukey’s method. A p value of < 0.05 was considered as significant.

4.6 Experiment 1: The Chair System’s Movement as Physical Motion

In Experiment 1, we informed to each subject that chair system’s movement was simple physical movement. We in-formed each subject that the objective of this experiment was to evaluate the effects of the simple physical movement to the subject’s emotion. Experiment 1 was conducted with 26 subjects (22 men and 4 women, aged 20.5 ± 1.7 years, mean ± S.D.). There were 7 subjects whose STAI scores after viewing the video stimulus were smaller than STAI score before viewing the video stimulus. We judged that the video stimulus did not invoke anxiety to these 7 subjects and excluded these 7 subjects from Experiment 1. There was one subject who reported that the subject thought that the chair system’s movement was generated by human’s breathing motion. Since the chair system’s movement should be recognized as simple physical movement in Experiment 1, we excluded this one subject from Experiment 1. In these way 26 − 7 – 1 = 18 subjects were analyzed in Experiment 1.

Result of Experiment 1

Figure 4 shows the results of STAI scores in Experiment 1. STAI score before viewing the video stimulus was 37.67 ± 7.0. STAI score after viewing the video stimulus and with Breath Chair Off was 49.7 ± 9.3. The STAI score after viewing the video stimulus and with Breath Chair On was 47.2 ± 11.7. The statistically significant difference was found between STAI score before viewing the video stimulus and STAI score after viewing the video stimulus based on one factor variance of analysis (F(2, 12) = 6.23, p < 0.001). The statistically significant difference was found between STAI score before viewing the video stimulus and STAI score with Breath Chair Off based on multiple comparison (p < 0.001). The statistically significant difference was found between STAI score before viewing the video stimulus and STAI score with Breath Chair On based on multiple comparison (p < 0.001).

Fig. 4.
figure 4

Results of STAI score in experiment 1 (n = 18, ***p < 0.001)

Full size image

The results for fingertip surface skin temperature are shown in Fig. 5 (210 s during the video) and in Fig. 6 (210 s immediately after the video). Two factor variance of analysis showed that video results had no significant difference. Results from immediately after the video showed main effect from presence or absence of Breath Chair On (F(1, 244) = 9.26, p < 0.05).

Fig. 5.
figure 5

Results of fingertip surface skin temperature during video in experiment 1 (n = 18)

Full size image
Fig. 6.
figure 6

Results of fingertip surface skin temperature after viewing the video in experiment 1(n = 18, *p < 0.05)

Full size image

4.7 Experiment 2: The Chair System’s Movement as Breathing Motion

In Experiment 2, we informed to each subject that the chair system’s movement was simulated by measuring human’s breathing motion. We informed each subject that the objective of this experiment was to evaluate the effects of the human’s breathing motion to the subject’s emotion. To let the subjects believe that the chair system’s movement was generated by measuring human’s breathing motion, we gave subjects wrong information that “there is another person in an adjacent room, who is comfortably resting, and this person’s breathing motion is remote sensed in real time and this measured motion generates the chair system’s movement” before Experiment 2.

Experiment 2 was conducted with 30 subjects (23 men and 7 women, aged 21.5 ± 1.6 years) who did not conduct Experiment 1. There were 6 subjects whose STAI scores after viewing the video stimulus were smaller than STAI score before viewing the video stimulus. We judged that the video stimulus did not invoke anxiety to these 6 subjects and excluded these 6 subjects from Experiment 1. There were 3 subjects who reported that the subjects thought that the chair system’s movement was not caused by human’s breathing motion. Since the chair system’s movement should be recognized as human’s breathing motion in Experiment 2, we excluded this one subject from Experiment 1. In these way 30 – 6 – 3 = 21 subjects were analyzed in Experiment 2.

Result of Experiment 2

Figure 7 shows the results of STAI scores in Experiment 2. STAI score before viewing the video stimulus was 36.2 ± 6.5. STAI score after viewing the video stimulus and with Breath Chair Off was 49.6 ± 9.7. The STAI score after viewing the video stimulus and with Breath Chair On was 44.7 ± 10.9. The statistically significant difference was found between STAI score before viewing the video stimulus and STAI score after viewing the video stimulus based on one factor variance of analysis (F(2, 40) = 25.3, p < 0.001). The statistically significant difference was found between STAI score before viewing the video stimulus and STAI score with Breath Chair Off based on multiple comparison (p < 0.001). The statistically significant difference was found between STAI score before viewing the video stimulus and STAI score with Breath Chair On based on multiple comparison (p < 0.001). The statistically significant difference was found between STAI score with Breath Chair On and STAI score with Breath Chair On based on multiple comparison (p < 0.05).

Fig. 7.
figure 7

Results of STAI score in experiment 2 (n = 21, *p < 0.05, ***p < 0.001)

Full size image

The results of fingertip surface skin temperature are shown in Fig. 8 (210 s when each subject was seeing the video stimulus) and in Fig. 9 (210 s immediately after each subject had shown the video stimulus). Based on two factor variance of analysis, there was no statistically significant difference during the video. After viewing the video, there was statistically significant difference between Breath Chair On and Breath Chair Off (F(1, 286) = 5.57, p < 0.05).

Fig. 8.
figure 8

Results of fingertip surface skin temperature during video in experiment 2 (n = 21)

Full size image
Fig. 9.
figure 9

Results of fingertip surface skin temperature after viewing the video in experiment 2 (n = 21, *p < 0.05)

Full size image

Among 21 subjects in total, only 2 subjects reported that the chair system’s movement disturbed the subjects to focus on the video at the discussion after each experiment, At the discussion after each experiment, we asked the question “Did you imagine the human’s persona whose breathing motion generates the chair system’s movement while you were in contact with the chair system’s movement?” to each subject. Among 21 subjects in total. 18 subjects answered that they did not imagine the human’s persona who generates the chair system’s movement. Other 3 subject thought that the experimenter’s breathing motion generated the chair system’s movement.

5 Discussion

Let us discuss the results of STAI state anxiety scale. In these experiments, we analyzed subjects whose fear or anxiety were invoked by the video stimulus (STAI scores of these subjects became higher by viewing the video stimulus under the condition with Breath Chair Off, in other words, the same condition as sitting in a regular chair). In Experiment 1, each subject recognized the chair system’s movements as simple physical movements, and in Experiment 2, each subject recognized the chair system’s movements as human’s breathing motions. There was statistically significant difference between the STAI score before viewing the video stimulus and the STAI score after viewing the video stimulus with Breath Chair Off. Thus video stimulus did invoke fear or anxiety to most subjects’ emotion. Although the STAI score with Breath Chair On was smaller than the STAI score with Breath Chair Off, there is a possibility that the fear or anxiety of the subjects did not reduced by the human’s breathing motion. There is a possibility that the subject’s attention to the video stimulus was distracted due to the chair system’s movement. However, in Experiment 1, the STAI score with Breath Chair On is bigger significantly than the STAI score before viewing the video stimulus and there was no significant difference between the STAI score with Breath Chair On and the STAI score with Breath Chair Off. Therefore, Breath Chair’s movement did not distract each subject’s attention to the video stimulus, and the video stimulus did invoke anxiety to each subject’s emotion. Furthermore, although in Experiment 1 (each subject recognized the chair system’s movements as simple physical movements), there was no significant difference between the STAI score with Breath Chair On and the STAI score with Breath Chair off, in Experiment 2 (each subject recognized the chair system’s movements as human’s breathing movements) the STAI score with Breath Chair On was significantly smaller than the STAI score with Breath Chair off. Therefore, simple physical movement cannot reduce the subject’s fear or anxiety, but human’s breathing motion can reduce the subject’s fear or anxiety. We want to conclude that since the subject thought the chair system’s movements as human’s breathing movements, the movements reduced the subject’s fear or anxiety. Nevertheless, there is a possibility that since the subjects was aware of the human whose breath generated the chair system’s movement in the adjacent room, this awareness distracted the subject’s attention to the video stimulus. However, at the discussion after the experiments, only 2 subjects reported that they were aware of the human in the adjacent room, and other 19 subjects were not aware of the human. These facts deny the above capability and we can conclude that since the subject recognized the chair system’s movement as human breath movement, it reduced the subject’s fear or anxiety.

Due to the stimulus which causes fear or anxiety, the autonomic nervous system shrinks peripheral blood vessels and the fingertip surface skin temperature decreases [31, 32]. The experiment of Kistler et al. [33] reported that fear or anxiety caused by video stimulus shrinks peripheral blood vessels and the fingertip surface skin temperature decreases. In our experiment also, we observed that the video stimulus caused fear or anxiety in the test subjects and decreased their fingertip surface skin temperature.

To analyze the fingertip surface skin temperature, the amount of change of it from the baseline is calculated, as we noticed in Sect. 4.5. This value becomes negative i.e. it is smaller than 0.

The amount of change of the fingertip surface skin temperature with Breath Chair On is larger than one with Breath Chair Off significantly in Experiment 2. The chair system’s movement might reduce fear and anxiety of each subject. There is another possibility that each subject’s fear and anxiety were reduced by the security feeling which was caused by the information that “there is another person in an adjacent room.” But the latter possibility is denied by the Experiment 1. The amount of change of the fingertip surface skin temperature with Breath Chair On is larger than one with Breath Chair Off significantly also in Experiment 1 in which the chair system’s movement was simple physical movement and there are no another person who gave security feeling from an adjacent room. Thus we claim that the chair system’s movement reduced the subject’s fear and anxiety. There was no statistically significant difference between the amount of change of the fingertip surface skin temperature with Breath Chair On and one with Breath Chair Off during viewing the video stimulus in Experiment 1 and 2. While each subject was viewing the video stimulus, it was invoking fear or anxiety to the subject and the effect of the chair system’s movement did not appear. After viewing the video stimulus, it stopped to invoke fear or anxiety to the subject and the effect of the chair system’s movement appeared.

By analyzing the results of STAI state anxiety scale, fingertip surface skin temperature, and the subjects’ reports after experiment, we claim that physical contact using simulated breath in our system Breath Chair reduced fear or anxiety.

Furthermore, when the subject was contact with Breath Chair’s movement as human’s breathing motion, the subject did not even imagine specific person whose breath generated the Breath Chair’s movement. This fact indicates that the simulated breath movement does not have to be caused by a related person who is familiar with the subject, and the simulated breath movement caused by an unrelated person may reduce the subject’s fear or anxiety.

6 Conclusion

We conducted experiments using a chair system “Breath Chair” which was developed by us [12]. We verified that the subject’s fear or anxiety could be reduced when the subject contact physically with the simulated breath which are caused by a per-son who is unrelated to the subject.

Starting with the field of tele-existence, there are attempts to support communication with specific people. However, there is little effort and study to support communication with unrelated person although the communication with unrelated person can reduce fear or anxiety. As a future outlook, we plan to develop and study systems about the positive effects caused by the coexistence of people who are not limited to those in close relationships such as friends, romantic partners, and family.