Showing papers presented at "International Conference on Social Robotics in 2014"

Perception of an Android Robot in Japan and Australia: A Cross-Cultural Comparison

Abstract: This paper reports the results from two experiments, conducted in Japan and Australia, to examine people’s perception and trust towards an android robot. Experimental results show that, in contrast to popular belief, Australian participants perceived the robot more positive than Japanese participants. This is the first study directly comparing human perception of a physically present android robot in two different countries.

Robots in Older People’s Homes to Improve Medication Adherence and Quality of Life: A Randomised Cross-Over Trial

Abstract: Healthcare robots are being developed to help older people maintain independence. This randomised cross-over trial aimed to investigate whether healthcare robots were acceptable and feasible and whether the robots could impact quality of life, depression and medication adherence. 29 older adults living in independent units within a retirement village were given robots in their homes for 6 weeks and had a non-robot 6-week control period, in a randomised order. The robots reminded people to take medication, provided memory games, entertainment, skype calls, and blood pressure measurement. The robots were found to be acceptable and feasible, and many participants described them as useful and as friends although not all comments were positive. There were relatively few problems with robot functions. The participants’ perceptions of the robots’ agency reduced over time. The robots had no significant impact on adherence, depression or quality of life. While the robots were feasible and acceptable, improvements in their reliability and functionality may increase their efficacy.

Activity Recognition for Natural Human Robot Interaction

Abstract: The ability to recognize human activities is necessary to facilitate natural interaction between humans and robots. While humans can distinguish between communicative actions and activities of daily living, robots cannot draw such inferences effectively. To allow intuitive human robot interaction, we propose the use of human-like stylized gestures as communicative actions and contrast them from conventional activities of daily living. We present a simple yet effective approach of modelling pose trajectories using directions traversed by human joints over the duration of an activity and represent the action as a histogram of direction vectors. The descriptor benefits from being computationally efficient as well as scale and speed invariant. In our evaluation, the descriptor returned state of the art classification accuracies using off the shelf classification algorithms on multiple datasets.

Robot Pressure: The Impact of Robot Eye Gaze and Lifelike Bodily Movements upon Decision-Making and Trust

Abstract: Between people, eye gaze and other forms of nonverbal communication can influence trust. We hypothesised similar effects would occur during human-robot interaction, predicting a humanoid robot’s eye gaze and lifelike bodily movements (eye tracking movements and simulated “breathing”) would increase participants’ likelihood of seeking and trusting the robot’s opinion in a cooperative visual tracking task. However, we instead found significant interactions between robot gaze and task difficulty, indicating that robot gaze had a positive impact upon trust for difficult decisions and a negative impact for easier decisions. Furthermore, a significant effect of robot gaze was found on task performance, with gaze improving participants’ performance on easy trials but hindering performance on difficult trials. Participants also responded significantly faster when the robot looked at them. Results suggest that robot gaze exerts “pressure” upon participants, causing audience effects similar to social facilitation and inhibition. Lifelike bodily movements had no significant effect upon participant behaviour.

Human robot interaction and fiction: A contradiction

Abstract: In this position paper a perspective on how movies and science fiction are currently shaping the design of robots is presented. This analysis includes both behaviour and embodiment in robots. We discuss popular movies that involve robots as characters in their storyline, and how people’s beliefs and expectations are affected by what they see in robot movies. A mismatch or contradiction emerges in what the robots of today can accomplish and what the movies portray. In order to overcome this mismatch we present design implications that may be of benefit to HRI designers.

A Sociological Contribution to Understanding the Use of Robots in Schools: The Thymio Robot

Abstract: The Thymio II robot was designed to be used by teachers in their classrooms for a wide range of activities and at all levels of the curriculum, from very young children to the end of high school. Although the educationally oriented design of this innovative robot was successful and made it possible to distribute more than 800 Thymio robots in schools with a large majority in the French-speaking part of Switzerland, it was not sufficient to significantly raise the number of teachers using robot technology in their teaching after three years of commercialization. After an introduction and a first section on the design of this educational robot, this paper presents some results of a sociological analysis of the benefits and blockages identified by teachers in using robots, or not, with their pupils.

Towards an Architecture for Knowledge Representation and Reasoning in Robotics

Abstract: This paper describes an architecture that combines the complementary strengths of probabilistic graphical models and declarative programming to enable robots to represent and reason with qualitative and quantitative descriptions of uncertainty and domain knowledge. An action language is used for the architecture’s low-level (LL) and high-level (HL) system descriptions, and the HL definition of recorded history is expanded to allow prioritized defaults. For any given objective, tentative plans created in the HL using commonsense reasoning are implemented in the LL using probabilistic algorithms, and the corresponding observations are added to the HL history. Tight coupling between the levels helps automate the selection of relevant variables and the generation of policies in the LL for each HL action, and supports reasoning with violation of defaults, noisy observations and unreliable actions in complex domains. The architecture is evaluated in simulation and on robots moving objects in indoor domains.

Ethical Dimensions of Human-Robot Interactions in the Care of Older People: Insights from 21 Focus Groups Convened in the UK, France and the Netherlands

Abstract: We briefly report the method and four findings of a large-scale qualitative study of potential users’ views on the ethical values that should govern the design and programming of social robots for older people. 21 focus groups were convened in the UK, France and the Netherlands. We present and briefly discuss our data on: 1) the contrasting attitudes of older people and formal and informal carers about how well technology might be received by older users; 2) views about healthcare professionals, informal and formal carers having access to private information about householders that has been collected by the robot; 3) the belief that robots could not, as well as should not, replace human contact because persuasion is regarded a uniquely human skill; and 4) differing perceptions of the role of the robot and how this was used to justify ethical opinions on robot behavior.

Impact of a Social Humanoid Robot as a Therapy Assistant in Children Cancer Treatment

Abstract: Treating cancer encompasses many invasive procedures that can be a source of distress in oncology patients Distress itself can be a major obstruction in the path of acceptance of treatment and the patient’s adaptation to it, thereby reducing its efficiency These distress symptoms have been found to be prevalent in children suffering from cancer, in a spectrum from mild to critical In the past years in response to this psychological suffering, researchers have proposed and tested several methods such as relaxation, hypnosis, desensitization, and distraction This paper propounds a new approach by exploring the effect of utilizing a humanoid robot as a therapy-assistive tool in dealing with pediatric distress Ten children, ages 6-10, diagnosed with cancer were randomly assigned into two groups of SRAT (5 kids) and psychotherapy (5 kids) at two specialized hospitals in Tehran A NAO robot was programmed and employed as a robotic assistant to a psychologist in the SRAT group to perform various scenarios in eight intervention sessions The promising results of this study in the level of anger, depression, and anxiety could render using social robots applicable in psychological interventions for children with cancer Results of this study shall be beneficial to psychologists, oncologists, and robot specialists

The Uncanny Valley: A Focus on Misaligned Cues

Abstract: Increasingly, humanoid robots and androids are easing into society for a wide variety of different uses. Previous research has shown that careful design of such robots is crucial as subtle flaws in their appearance, vocals and movement can give rise to feelings of unease in those interacting with them. Recently, the Bayesian model for the uncanny has suggested that conflicting or misaligned cues at category boundaries may be the main attributing factor of this phenomenon. The results from this study imply that this is indeed the case and serve as empirical evidence for the Bayesian theory.

Group Comfortability When a Robot Approaches

Abstract: This paper investigates the level of comfort in people with different robot approach paths. While engaged in a shared task, 45 pairs of participants were approached by a robot from eight different directions and asked to rate their level of comfort. Results show that comfortability patterns of individuals in pairs is different to lone individuals when they are approached by a robot. This in turn influences how comfortable a group is with different robot approach paths.

Investigating the Effect of a Robotic Tutor on Learner Perception of Skill Based Feedback

Abstract: In this paper we investigate the effect of different embodiments on perception of a skill based feedback (a basic open learner model) with a robotic tutor. We describe a study with fifty-one 11-13 year old learners. Each learner carries out a geography based activity on a touch table. A real time model of the learner’s skill levels is built based on the learner’s interaction with the activity. We explore three conditions where the contents of this learner model is fed back to the learner with different levels of embodiment: (1) Full embodiment, where skill levels are presented and explained solely by a robot; (2) Mixed embodiment, where skill levels are presented on a screen with explanation by a robot; and (3) No embodiment, where skill levels and explanation are presented on a screen with no robot. The findings suggest that embodiment can increase enjoyment, understanding, and trust in explanations of an open learner model.

To Beep or Not to Beep is Not the Whole Question

Abstract: In this paper, we address social effects of different mechanisms by means of which a robot can signal a person that it wants to pass. In the situation investigated, the robot attempts to pass by a busy, naive participant who is blocking the way for the robot. The robot is a relatively large service robot, the Care-o-bot. Since speech melody has been found to fulfill social functions in human interactions, we investigate whether there is a difference in perceived politeness of the robot if the robot uses a beep sequence with rising versus with falling intonation, in comparison with no acoustic signal at all. The results of the experimental study (n=49) shows that approaching the person with a beep makes people more comfortable than without any sound, and that rising intonation contours make people feel more at ease than falling contours, especially women, who rate the robot that uses rising intonation contours as friendlier and warmer. The exact form of robot output thus matters.

An Approach to Socially Compliant Leader Following for Mobile Robots

Abstract: Mobile robots are envisioned to provide more and more services in a shared environment with humans. A wide range of such tasks demand that the robot follows a human leader, including robotic co-workers in factories, autonomous shopping carts or robotic wheelchairs that autonomously navigate next to an accompanying pedestrian. Many authors proposed follow-the-leader approaches for mobile robots, which have also been applied to the problem of following pedestrians. Most of these approaches use local control methods to keep the robot at the desired position. However, they typically do not incorporate information about the natural navigation behavior of humans, who strongly interact with their environment. In this paper, we propose a learned, predictive model of interactive navigation behavior that enables a mobile robot to predict the trajectory of its leader and to compute a far-sighted plan that keeps the robot at its desired relative position. Extensive experiments in simulation as well as with a real robotic wheelchair suggest that our method outperforms state-of-the-art methods for following a human leader in wide variety of situations.

Classical Conditioning in Social Robots

Abstract: Classical conditioning is important in humans to learn and predict events in terms of associations between stimuli and to produce responses based on these associations. Social robots that have a classical conditioning skill like humans will have an advantage to interact with people more naturally, socially and effectively. In this paper, we present a novel classical conditioning mechanism and describe its implementation in ASMO cognitive architecture. The capability of this mechanism is demonstrated in the Smokey robot companion experiment. Results show that Smokey can associate stimuli and predict events in its surroundings. ASMO’s classical conditioning mechanism can be used in social robots to adapt to the environment and to improve the robots’ performances.

Using a gaze-cueing paradigm to examine social cognitive mechanisms of individuals with autism observing robot and human faces

Abstract: This paper reports a study in which we investigated whether individuals with autism spectrum disorder (ASD) are more likely to follow gaze of a robot than of a human. By gaze following, we refer to one of the most fundamental mechanisms of social cognition, i.e., orienting attention to where others look. Individuals with ASD sometimes display reduced ability to follow gaze [1] or read out intentions from gaze direction [2]. However, as they are in general well responding to robots [3], we reasoned that they might be more likely to follow gaze of robots, relative to humans. We used a version of a gaze cueing paradigm [4, 5] and recruited 18 participants diagnosed with ASD. Participants were observing a human or a robot face and their task was to discriminate a target presented either at the side validly cued by the gaze of the human or robot; or at the opposite side. We observed typical validity effects: faster reaction times (RTs) to validly cued targets, relative to invalidly cued targets. However, and most importantly, the validity effect was larger and significant for the robot faces, as compared to the human faces, where the validity effect did not reach significance. This shows that individuals with ASD are more likely to follow gaze of robots, relative to humans, suggesting that the success of robots in involving individuals with ASD in interactions might be due to a very fundamental mechanism of social cognition. Our present results can also provide avenues for future training programs for individuals with ASD.

Representation and Execution of Social Plans through Human-Robot Collaboration

Abstract: The use of robots in people daily life and, accordingly, the requirement for a robot to behave in a socially acceptable way are getting more and more attention. However, although many progresses have been done in the last years, robots still have many limitations when they are required to share the environment with humans.

Development of Brain Training Games for a Healthcare Service Robot for Older People

Abstract: Many people suffer from memory loss as they get older. This may lead to severe memory conditions, where everyday communication and activities for a person become much more difficult and independent living is a challenge. Brain training is one therapeutic method for people who are concerned about brain function decline. We previously found that brain training games are helpful and enjoyable for older people. In this paper, we describe the development of new computer-based brain training games based on paper-based brain training exercises created by experts on brain function. We develop four games targeting specific areas of memory, prospective memory, face recognition skills, verbal memory, and short-term memory. These games are deployed on our healthcare service robot, which is used in an individual home environment. We report a usability study with older adults to evaluate our brain training games on a healthcare service robot. Results show the games are usable and people responded positively about them, and some improvements were identified for future development.

Social Robotics through an Anticipatory Governance Lens

Abstract: Social Robotics is an emerging field, with many applications envisioned. Scientific and technological advancements constantly impact humans on the individual and societal level. Therefore one question increasingly debated is how to anticipate the impact of a given envisioned, emerging or new scientific or technological development and how to govern the emergence of scientific and technological advancements. Anticipatory governance has as a goal to discuss potential issues arising at the ground level of the emergence of a given scientific and technological product. Our study investigated a) the visibility of the anticipatory governance concept within the social robotic discourse and b) the implication of anticipatory governance for the social robotics field through the lens of a social robot design process and key documents from the UNESCO/ICSU 1999 World Conference on Sciences the lens. Our findings suggest that a) anticipatory governance is not a concept established within the social robotics fields so far; b) that social robotics as specific field is not engaged with within the anticipatory governance field and c) that many professional and academic fields are not yet involved in the social robotics discourse as aren’t many non-academic stakeholders. We posit that anticipatory governance can strengthen the social robotics field.

Skeleton Tracking Based Complex Human Activity Recognition Using Kinect Camera

Abstract: This paper presents a new and efficient algorithm for complex human activity recognition using depth videos recorded from a single Microsoft Kinect camera. The algorithm has been implemented on videos recorded from Kinect camera in OpenNI video file format (.oni). OpenNI file format provides a combined video with both RGB and depth information. An OpenNI specific dataset of such videos has been created containing 200 videos of 8 different activities being performed by different individuals. This dataset should serve as a reference for future research involving OpenNI skeleton tracker. The algorithm is based on skeleton tracking using state of the art OpenNI skeleton tracker. Various joints and body parts in human skeleton have been tracked and the selection of these joints is made based on the nature of the activity being performed. The change in position of the selected joints and body parts during the activity has been used to construct feature vectors for each activity. Support vector machine (SVM) multi-class classifier has been used to classify and recognize the activities being performed. Experimental results show the algorithm is able to successfully classify the set of activities irrespective of the individual performing the activities and the position of the individual in front of the camera.

Motion-Oriented Attention for a Social Gaze Robot Behavior

Abstract: Various studies have shown that human visual attention is generally attracted by motion in the field of view. In order to embody this kind of social behavior in a robot, its gaze should focus on key points in its environment, such as objects or humans moving. In this paper, we have developed a social natural attention system and we explore the perception of people while interacting with a robot in three different situations: one where the robot has a totally random gaze behavior, one where its gaze is fixed on the person in the interaction, and one where its gaze behavior adapts to the motion-based environmental context. We conducted an online survey and an on-site experiment with the Meka robot so as to evaluate people’s perception towards these three types of gaze. Our results show that motion-oriented gaze can help to make the robot more engaging and more natural to people.

Using Robots to Modify Demanding or Impolite Behavior of Older People

Abstract: As part of a large scale qualitative study (conducted in France, the UK and the Netherlands) of potential users’ views on the ethical values that should govern the design and programming of social robots for older people, we elicited responses to a scenario where a robot is programmed to modify an older person’s rude behavior. Participants’ responses ranged from outright disagreement with robotized efforts to change characteristic behavior, to approval as a means to an end. We discuss these views against the background of respect for autonomy, the differences and similarities between robot and human carers, and behavior modification in the context of rehabilitation, where the ‘no gain without pain’ principle is commonly used to justify what would otherwise seem callous. We conclude that such programming may be acceptable in the context of the rehabilitation and promotion of the independence of older people.

Robots and the Division of Healthcare Responsibilities in the Homes of Older People

Abstract: This paper briefly describes the method of a qualitative study, which used focus groups to elicit the views of older people and formal and informal carers of older people on the ethical issues surrounding the introduction of social robots into the homes of older people. We then go on to sketch some of the tensions and conflicts that can arise between formal carers, informal carers, and older people when trying to negotiate the task of dividing care responsibilities, and describe how the introduction of robots may exacerbate, or ease, these tensions. Data from the qualitative study is used to indicate where participants acknowledged, identified and discussed these issues.

Real-Time Gender Based Behavior System for Human-Robot Interaction

Abstract: This work introduces a real-time system able to lead humanoid robot behavior depending on the gender of the interacting person. It exploits Aldebaran NAO humanoid robot view capabilities by applying a gender prediction algorithm based on the face analysis. The system can also manage multiple persons at the same time, recognizing if the group is composed by men, women or is a mixed one and, in the latter case, to know the exact number of males and females, customizing its response in each case. The system can allow for applications of human-robot interaction requiring an high level of realism, like rehabilitation or artificial intelligence.

Social Robots in Postural Education: A New Approach to Address Body Consciousness in ASD Children

Abstract: Autism Spectrum Disorders (ASD) represent one of the most prevalent developmental disorders among children with different level of impairments in social relationships, communication and imagination. In addition, impaired movement is also observed in individuals with ASD and recent studies consider this factor as a limitation for fully engagement in the social environment. In the present work, we propose a new approach to promote postural education in autistic children with the involvement of a humanoid social robot and the therapist in a triadic interaction environment to better understand their motor development and body consciousness.

Integrating Reinforcement Learning and Declarative Programming to Learn Causal Laws in Dynamic Domains

Abstract: Robots deployed to assist and collaborate with humans in complex domains need the ability to represent and reason with incomplete domain knowledge, and to learn from minimal feedback obtained from non-expert human participants. This paper presents an architecture that combines the complementary strengths of Reinforcement Learning (RL) and declarative programming to support such commonsense reasoning and incremental learning of the rules governing the domain dynamics. Answer Set Prolog (ASP), a declarative language, is used to represent domain knowledge. The robot’s current beliefs, obtained by inference in the ASP program, are used to formulate the task of learning previously unknown domain rules as an RL problem. The learned rules are, in turn, encoded in the ASP program and used to plan action sequences for subsequent tasks. The architecture is illustrated and evaluated in the context of a simulated robot that plans action sequences to arrange tabletop objects in desired configurations.

Application of Fuzzy Techniques in Human-Robot Interaction - A Review

Abstract: Targeting research challenges in Socially Assistive Robotics (SAR), this paper provides a review of previous work that describe robot or non-robot systems that use fuzzy logic to infer high-level human intention or activities. In comparison to statistical and probabilistic approaches which are very popular in SAR and Human-Robot Interaction (HRI), this review focuses on fuzzy logic-based systems. As fuzzy logic has already been widely used in almost all research areas in robotics, this review does not consider systems that uses fuzzy logic for sensing, modelling or planning tasks except for inferencing or reasoning tasks. From this review, it was found minimal research has been done in this special research niche and is deemed to gain more attention as the research communities shifts from sensing toward modelling and inferencing in the loop of Sense-Model-Plan-Act or Sense-Plan-Act.

Motivational Emotion Generation and Behavior Selection based on Emotional Experience for Social Robots

Abstract: As the number of social and psychological service robots increase, many robots which have emotions have been developed. In order to implement an emotion generation model, the authors of this paper suggest to employ the unified motivation theory which is proposed by Higgins. In addition, a behavior selection model based on emotional experience is suggested. A joint attention scenario is suggested in which a human and a robot play a 2048 game together with a follow-up interaction. A preliminary pilot test was conducted to examine the proposed method and it performed properly as intended. Finally, further experiment plans and expectations are discussed.

Sociable Dining Table: Incremental Meaning Acquisition Based on Mutual Adaptation Process

Abstract: Our main goal is to explore how social interaction can evolve incrementally and be materialized in a protocol of communication. We intend to study how the human establishes a protocol of communication in a context that requires mutual adaptation. Sociable Dining Table (SDT) integrates a dish robot put on the table and behaves according to the knocks that the human emits. To achieve our goal, we conducted two experiments: a human-controller experiment (Wizard-of-Oz) and a human-robot interaction (HRI) experiment. The aim of the first experiment is to understand how people are building a protocol of communication. We suggest an actor-critic architecture that simulates in an open ended way the adaptive behavior that we have seen in the first experiment. We show in a human-robot interaction (HRI) experiment that our method enables the adaptation to the individual preferences in order to get a personalized protocol of communication.

ROBOMO: Towards an Accompanying Mobile Robot

Abstract: Many of the most daily uses of robots require them to work alongside users as cooperative and socially adaptive partners. To provide the human with the better suited assistance at a convenient time, a robot must assimilate the user’s behaviors and afford an adaptive response within the context of the interaction they share. We try to understand how a robot communication that is based on inarticulate sounds and iconic gestures is capable to help on the establishment of the attachment process and can enhance the social bonding between the human and our accompanying mobile robot (ROBOMO). In this paper, we draw on inarticulate sound and iconic gestures in order to design our robot and ground the attachment process. We showed that using simple inarticulate sounds and iconic gestures, the attachment process can evolve incrementally which significantly helped to acquire the meaning of the robot’s behaviors.