Social Robotics

Social acceptance of robots in public space implicitly requires a user-centered approach in the development phase of a robotic system. One of the most popular tools used by interaction designers in the field of Human-Computer Interaction (HCI) are the persona-based scenarios. Recently, they find their way more and more into HRI research. This paper describes how specifically designed human-human studies can support the modeling process of interaction scenarios in HRI. As part of the Interactive Urban Robot (IURO) project, we analyzed three human-human studies to better understand human selection process of interaction partners and an approached person’s subsequent actions in itinerary request situations. This empirical data was used to create personas and scenarios for HRI in public space.

In this paper, we propose a MAWARI-based social interface as an interactive social medium to broadcast information (e.g. news, etc.). The interface consists of three sociable creatures (MAWARIs) and is designed with minimalism designing concepts. MAWARI is a small scale robot that has only body gestures to express (or interact) its attractive social cues. This helps to reduce the observing workload of the user, because during the conversation the creature just uses vocal interactions mixed with attractive body gestures. The proposed interface is capable of behaving in two kinds of states: behavior as a passive social interface and also as an interactive social interface in order to reduce the conversation workload of the participant.

Personal robots that share the same space with humans need to be socially acceptable and effective as they interact with people. In this paper we focus our attention on the definition of a behaviour-based robotic architecture that, (1) allows the robot to navigate safely in a cluttered and dynamically changing domestic environment and (2) encodes embodied non-verbal interactions: the robot respects the user’s personal space by choosing the appropriate distance and direction of approach. The model of the personal space is derived from a human-robot psycho-physical study and it is described in a convenient mathematical form. The robot’s target location is dynamically inferred through the solution of a Bayesian filtering problem. The validation of the overall behavioural architecture shows that the robot is able to exhibit appropriate proxemic behaviour.

Previous work has shown that gestural behaviors affect anthropomorphic inferences about artificial communicators such as virtual agents. In an experiment with a humanoid robot, we investigated to what extent gesture would affect anthropomorphic inferences about the robot. Particularly, we examined the effects of the robot’s hand and arm gestures on the attribution of typically human traits, likability of the robot, shared reality, and future contact intentions after interacting with the robot. For this, we manipulated the non-verbal behaviors of the humanoid robot in three experimental conditions: (1) no gesture, (2) congruent gesture, and (3) incongruent gesture. We hypothesized higher ratings on all dependent measures in the two gesture (vs. no gesture) conditions. The results confirm our predictions: when the robot used gestures during interaction, it was anthropomorphized more, participants perceived it as more likable, reported greater shared reality with it, and showed increased future contact intentions than when the robot gave instructions without using gestures. Surprisingly, this effect was particularly pronounced when the robot’s gestures were partly incongruent with speech. These findings show that communicative non-verbal behaviors in robotic systems affect both anthropomorphic perceptions and the mental models humans form of a humanoid robot during interaction.

Robot Theater is a fairly new arena for researching Human Robot Interaction, however, in surveying research already conducted, we have identified eight lessons from Robot Theater that inform the design of social robots today. As an interdisciplinary field, we include examples spanning robotics researchers, acting theorists, cognitive neuroscientists, behavioral psychologists and dramaturgy literature. Lessons learned include (1) the importance of intentionality in action; (2)(3)(4) the relationship between embodiment, gesture, and emotional expression; (5) the bipolar sociability categorization between machine and agent; (6) the power of interaction partners to shape robot attributions; (7) the role of audience acknowledgement and feedback; (8) the power of humor to enhance interaction. Robotics has had a long history with the field of entertainment; even the word ‘robot’ comes from the 1921 Czech play ‘R.U.R.’ - we look forward to rigorous and continued research and cross-pollination between these domains.

In this work, we discuss a set of metrics for analyzing human spatial behavior (proxemics) motivated by work in the social sciences. Specifically, we investigate individual, attentional, interpersonal, and physiological factors that contribute to social spacing. We demonstrate the feasibility of autonomous real-time annotation of these spatial features during multi-person social encounters. We utilize sensor suites that are non-invasive to participants, are readily deployable in a variety of environments (ranging from an instrumented workspace to a mobile robot platform), and do not interfere with the social interaction itself. Finally, we provide a discussion of the impact of these metrics and their utility in autonomous socially interactive systems.

Previous results show that adults are able to interpret different key poses displayed by the robot and also that changing the head position affects the expressiveness of the key poses in a consistent way. Moving the head down leads to decreased arousal (the level of energy), valence (positive or negative) and stance (approaching or avoiding) whereas moving the head up produces an increase along these dimensions [1]. Hence, changing the head position during an interaction should send intuitive signals which could be used during an interaction. The ALIZ-E target group are children between the age of 8 and 11. Existing results suggest that they would be able to interpret human emotional body language [2, 3].

Based on these results, an experiment was conducted to test whether the results of [1] can be applied to children. If yes body postures and head position could be used to convey emotions during an interaction.

Social agency theory suggests that when an (artificial) agent combines persuasive strategies, its persuasive power increases. Therefore, we investigated whether a robot that uses two persuasive strategies is more persuasive than a robot that uses only one. Because in human face-to-face persuasion two crucial persuasive strategies are gazing and gestures, the current research investigated the combined and individual contribution of gestures and gazing on the persuasiveness of a storytelling robot. A robot told a persuasive story about the aversive consequences of lying to 48 participants. The robot used persuasive gestures (or not) and gazing (or not) to accompany this persuasive story. We assessed persuasiveness by asking participants to evaluate the lying individual in the story told by the robot. Results indicated that only gazing independently led to increased persuasiveness. Using persuasive gestures only led to increased persuasiveness when the robot combined it with (the persuasive strategy of) gazing. Without gazing, using persuasive gestures diminished robot persuasiveness. The implications of the current findings for theory and design of persuasive robots are discussed.

Increasingly, people will be exposed to social robots. In order to inform the design of behaviors for robots that share domestic and public spaces with humans, it is important to know what robot behavior is considered as ‘normal’ by human users. The work reported in this paper stems from the premise that what would be perceived as socially normative behavior for robots may differ from what is considered socially normative for humans. This paper details the development of a set of measures, BEHAVE, for assessing user responses to a robot’s behavior using both attitudinal and physical responses. To test the validity and reliability of the BEHAVE set of measures, a human robot interaction experiment was conducted in which a robot invaded the personal space of a participant. Based on the results from this evaluation, a final set of BEHAVE measures was developed.

In this paper we present a study of how touch may be used as a way of inducing trust in the meeting with a teleoperated android, the Geminoid-DK. The use of haptics with Gestalt-based ’Balance Theory’ is used as a persuasive design of the interaction between human and robot. Balance theory includes both cognition and affect as factors influencing motivated behavior, subsequently effecting judgments of trustworthiness. Haptic interactions have both conscious and unconscious implications for the receiver in this interaction, which we demonstrate by an experimental set-up, as well as through questionnaires and interview.

With the development of artificially intelligent humanoid consorts, robotics is venturing into a realm of legal liability that has traditionally governed social interactions between humans and other humans, rather than interactions between humans and machines. How can and should legal systems deal with the problems that arise in regulated human interpersonal and sexual relationships when there is an AI sex doll in the mix? Heart balm torts, traditionally used to hold a third party paramour civilly liable for the dissolution of a protected relationship, provide a potential answer. Finding an appropriate entity to be liable will be problematic, though robot producers and the AI entity itself could both be potential defendants in a heart balm case. Producers may be able to limit liability if they can incorporate the experience of heartbreak and compassion into their creations.

This paper reports findings from an open-ended survey on attitudes towards humanoid robots collected from samples in the United Kingdom and Japan. 335 participants were asked how they felt about humanoid robots becoming widespread in society and what tasks they wanted humanoid robots to perform. While the UK sample was overall less negative towards humanoid robots than their Japanese counterparts, the UK sample did not want robots to perform tasks that required capabilities deemed as human qualities, such as empathy, caring, or independent decision making.

As robots are now being widely used as educational aids and assistants, it is crucial to understand the effects of robotic teaching assistants in classroom and how attraction and acceptance towards the robot are shaped. A 2 (type of instructor: human vs. robot) x 3 (feedback style: positive vs. negative vs. neutral) between-subjects experiment with six conditions was conducted to examine the effects of a robot instructor in classroom and the instructor’s feedbacks on students’ attraction and acceptance towards the given feedback. Results showed that feedback from a human instructor were more acceptable than feedback from a robot instructor. Students in the robot-instructor condition showed greater attraction towards the instructor when received a positive feedback, whereas students in the human-instructor condition did not report any difference in their attraction towards the instructor due to the feedback style. Both implications and limitations of the present study as well as guidelines for future research are discussed.

In this paper, we introduce a new form of an English education system that utilizes a teleoperated robot controlled by a native human teacher in a remote site. By providing a unique operation interface that incorporates non-contact vision recognition technologies, a native teacher easily and naturally controls the robot from a distance while providing lectures. Due to its effective representation of a human teacher in a robot, students show great interest on the robot and feel comfortable learning English with the teleoperated robot. In a real field pilot study, the participated elementary students have achieved good improvements on standardized tests after the study, which shows the effectiveness of the teleoperated robot system.

Based on a ‘learning by playing’ concept, a basic arithmetic learning task was extended with an engaging game to achieve long-term educational interaction for children. Personalization was added to this learning task, to further support the child’s motivation and success in learning. In an experiment, twenty children (aged 9-10) interacted three times, spread over days, with a robot using the combined imitation and arithmetic game to test this support. Two versions of the robot were implemented. In one implementation the complexity of the arithmetic progressed towards a predefined group target. In the other version the assignments increased in complexity until a personal end level was reached. A subsequent free-choice period showed that children’s motivation to play (and learn) was high, particularly when the game progressed to a personal target. Furthermore results show that most children in the last condition reach higher levels compared to the predefined group level.

The care robot of the future should be able to navigate in domestic environments and perform meaningful tasks. Presumably, a robot that moves and interacts more intelligently gains more trust, is liked more and appears more humanlike. Here we test in three scenarios of differing urgency whether anticipatory walking behaviour of a robot is appreciated as more intelligent and whether this results in a more positive attitude towards the robot. We find no effect of walking behaviour and a main effect of urgency of the scenarios on perceived intelligence and on appropriateness. We interpret these results as that the type of interaction determines perceived intelligence and the attitude towards robots, but the degree of anticipation has no significant effect.

Researchers have shown it is possible to develop robots that can produce recognizable emotional facial expressions [1, 2]. However, although human emotional expressions are known to be influenced by the surrounding context [7], there has been little research into the effect of context on the recognition of robot emotional expressions. The experiment reported here demonstrates that classical music can affect judgments of a robot’s emotional facial expressions. Different judgments were made depending on whether the music was emotionally congruent or incongruent with the robot’s expressions. A robot head produced sequences of expressions that were designed to demonstrate positive or negative emotions. The expressions were more likely to be recognized as intended when they occurred with music of a similar valence. Interestingly, it was observed that the robot face also influenced judgments about the classical music. Design implications for believable emotional robots are drawn.

This paper presents the analysis of human nonverbal responses and betting decision in terms of bluffing and the comparison between human-human poker game and human-robot poker game. According to a given situation manipulated with the card hand strength, we explored how different the participants change bluff decision between strong hand and weak hand and how different the participants use a bluff between a human-human and human-robot poker game. Furthermore, we analyzed the significant correlation between the card hand strength and human behaviors and obtained the regression model to predict the card hand strength from nonverbal behaviors and bluff decision.

Humans take different roles when they work together on a common task. But how do humans react to different roles of a robot in a human-robot interaction scenario? In this publication, we present a user evaluation, in which naïve participants work together with a robot on a common construction task. The robot is able to take different roles in the interaction: one group of the experiment participants worked with the robot in the instructive role, in which the robot first instructs the user how to proceed with the construction and then supports the user by handing over building pieces. The other group of participants used the robot in its supportive role, in which the robot hands over assembly pieces to the human that fit to the current progress of the assembly plan and only gives instructions when necessary. The results of the experiment show that the users do not prefer one of the two roles of the robot, but take the counterpart to the robot’s role and adjust their own behaviour according to the robot’s actions. This is revealed by the objective data that we collected as well as by the subjective answers of the experiment participants to a user questionnaire. The data suggests that the most influential factors for user satisfaction are the number of times the users picked up a building piece without getting an explicit instruction by the robot and the number of utterances the users made themselves. While the number of pickup actions had a positive or negative influence, depending on the role the users took, the number of own utterances always had a strong negative influence on the user’s satisfaction.

The study presented in this paper examined people’s perception of domestic service robots by means of an ethnographic study. We investigated initial reactions of nine households who lived with a Roomba vacuum cleaner robot over a two week period. To explore people’s attitude and how it changed over time, we used a recurring questionnaire that was filled at three different times, integrated in 18 semi-structured qualitative interviews. Our findings suggest that being part of a specific household has an impact how each individual household member perceives the robot. We interpret that, even though individual experiences with the robot might differ from one other, a household shares a specific opinion about the robot. Moreover our findings also indicate that how people perceived Roomba did not change drastically over the two week period.

RoboCup@Home is the largest benchmarking effort for domestic service robots. The benchmarking is in the form of a competition, with several yearly local competitions and an international one. Every year the tests become more complex, depending on the results of the previous years. In the past four years the focus has been on benchmarking physical aspects of the robots, such as manipulation, recognizing people and human-robot interaction. In 2010, for the first time, there is a test which is targeted at the mental cognitive capabilities of the robot. In order to guarantee scientific quality of the proposed solutions and effective integration in a fully working system, all the tests include different capabilities and change every year. This novel feature of RoboCup@Home benchmarking raises the question of: How can effective benchmark tests be defined and at the same time measure the progress over many years? In this paper we present the methodology applied in and results of RoboCup@Home for measuring the effectiveness of benchmarking service robots through competitions and present a new integrated test for benchmarking the cognitive abilities of a robot.

Social embodied agents may mitigate moments of apathy and confusion that older adults can experience at home. Based on a literature study, use cases, requirements and claims were specified. In an experiment with 29 older adults (aged 70+), it was studied to what extent a virtual agent and three robots (i.e., the Nao, iCat and Nabaztag) provide a platform to support these use cases, requirements and claims. Participants seemed to evaluate the agents mainly in terms of three generic components: the perceived level of realism, intellectuality and friendliness. A more serious and agreeable appearance improved the appreciation of the agent’s actions. Especially facial realism appeared to be important for trust, social presence, perceived sociability and perceived enjoyment.