Embodied artificial agents for understanding human social cognition

Beyond human-likeness: Socialness is more influential when attributing mental states to robots

We sought to replicate and expand previous work showing that the more human-like a robot appears, the more willing people are to attribute mind-like capabilities and socially engage with it. Forty-two participants played games against a human, a humanoid robot, a mechanoid robot, and a computer algorithm while undergoing functional neuroimaging. We confirmed that the more human-like the agent, the more participants attributed a mind to them. However, exploratory analyses revealed that the perceived socialness of an agent appeared to be as, if not more, important for mind attribution. Our findings suggest top-down knowledge cues may be equally or possibly more influential than bottom-up stimulus cues when exploring mind attribution in non-human agents. While further work is now required to test this hypothesis directly, these preliminary findings hold important implications for robotic design and to understand and test the flexibility of human social cognition when people engage with artificial agents.

Psychophysiological responses to eye contact with a humanoid robot: Impact of perceived intentionality

Eye contact with a social robot has been shown to elicit similar psychophysiological responses to eye contact with another human. However, it is becoming increasingly clear that the attention- and affect-related psychophysiological responses differentiate between direct (toward the observer) and averted gaze mainly when viewing embodied faces that are capable of social interaction, whereas pictorial or pre-recorded stimuli have no such capability. It has been suggested that genuine eye contact, as indicated by the differential psychophysiological responses to direct and averted gaze, requires a feeling of being watched by another mind. Therefore, we measured event-related potentials (N170 and frontal P300) with EEG, facial electromyography, skin conductance, and heart rate deceleration responses to seeing a humanoid robot's direct versus averted gaze, while manipulating the impression of the robot's intentionality. The results showed that the N170 and the facial zygomatic responses were greater to direct than to averted gaze of the robot, and independent of the robot's intentionality, whereas the frontal P300 responses were more positive to direct than to averted gaze only when the robot appeared intentional. The study provides further evidence that the gaze behavior of a social robot elicits attentional and affective responses and adds that the robot's seemingly autonomous social behavior plays an important role in eliciting higher-level socio-cognitive processing.

Attention allocation in complementary joint action: How joint goals affect spatial orienting

When acting jointly, individuals often attend and respond to the same object or spatial location in complementary ways (e.g., when passing a mug, one person grasps its handle with a precision grip; the other receives it with a whole-hand grip). At the same time, the spatial relation between individuals' actions affects attentional orienting: one is slower to attend and respond to locations another person previously acted upon than to alternate locations ("social inhibition of return", social IOR). Achieving joint goals (e.g., passing a mug), however, often requires complementary return responses to a co-actor's previous location. This raises the question of whether attentional orienting, and hence the social IOR, is affected by the (joint) goal our actions are directed at. The present study addresses this question. Participants responded to cued locations on a computer screen, taking turns with a virtual co-actor. They pursued either an individual goal or performed complementary actions with the co-actor, in pursuit of a joint goal. Four experiments showed that the social IOR was significantly modulated when participant and co-actor pursued a joint goal. This suggests that attentional orienting is affected not only by the spatial but also by the social relation between two agents' actions. Our findings thus extend research on interpersonal perception-action effects, showing that the way another agent's perceived action shapes our own depends on whether we share a joint goal with that agent.

Biological postural oscillations during facial expression of pain in virtual characters modulate early and late ERP components associated with empathy: A pilot study

There is a surge in the use of virtual characters in cognitive sciences. However, their behavioural realism remains to be perfected in order to trigger more spontaneous and socially expected reactions in users. It was recently shown that biological postural oscillations (idle motion) were a key ingredient to enhance the empathic response to its facial pain expression. The objective of this study was to examine, using electroencephalography, whether idle motion would modulate the neural response associated with empathy when viewing a pain-expressing virtual character. Twenty healthy young adults were shown video clips of a virtual character displaying a facial expression of pain while its body was either static (Still condition) or animated with pre-recorded human postural oscillations (Idle condition). Participants rated the virtual human's facial expression of pain as significantly more intense in the Idle condition compared to the Still condition. Both the early (N2-N3) and the late (rLPP) event-related potentials (ERPs) associated with distinct dimensions of empathy, affective resonance and perspective-taking, respectively, were greater in the Idle condition compared to the Still condition. These findings confirm the potential of idle motion to increase empathy for pain expressed by virtual characters. They are discussed in line with contemporary empathy models in relation to human-machine interactions.

Virtual agents and risk-taking behavior in adolescence: the twofold nature of nudging

Peer pressure can influence risk-taking behavior and it is particularly felt during adolescence. With artificial intelligence (AI) increasingly present in a range of everyday human contexts, including virtual environments, it is important to examine whether AI can have an impact on human's decision making processes and behavior. By using the balloon analogue risk task (BART) evaluating propensity to take risk, in this study 113 adolescents' risk-taking behavior was measured when playing alone and in the presence of either a robot avatar or human avatar. In the avatar conditions, participants performed the BART while the avatars either (1) verbally incited risk-taking or (2) discouraged risk-taking (experimental tasks). Risk-taking behavior in the BART was assessed in terms of total number of pumps, gain and explosions. Tendency to impulsivity was also evaluated, as well as the effects of age and gender on risky behavior. The main finding showed a significant effect of both avatars on risk-taking tendency, with riskier behavior during incitement than discouragement conditions, the latter being also substantially different from the playing-alone condition. The results of this study open up new questions in a very sensitive and timely topic and offer various insights into the effect of nudging on adolescents' behavior in virtual contexts.

The promise and peril of interactive embodied agents for studying non-verbal communication: a machine learning perspective

In face-to-face interactions, parties rapidly react and adapt to each other's words, movements and expressions. Any science of face-to-face interaction must develop approaches to hypothesize and rigorously test mechanisms that explain such interdependent behaviour. Yet conventional experimental designs often sacrifice interactivity to establish experimental control. Interactive virtual and robotic agents have been offered as a way to study true interactivity while enforcing a measure of experimental control by allowing participants to interact with realistic but carefully controlled partners. But as researchers increasingly turn to machine learning to add realism to such agents, they may unintentionally distort the very interactivity they seek to illuminate, particularly when investigating the role of non-verbal signals such as emotion or active-listening behaviours. Here I discuss some of the methodological challenges that may arise when machine learning is used to model the behaviour of interaction partners. By articulating and explicitly considering these commitments, researchers can transform 'unintentional distortions' into valuable methodological tools that yield new insights and better contextualize existing experimental findings that rely on learning technology. This article is part of a discussion meeting issue 'Face2face: advancing the science of social interaction'.

Modulations of one's sense of agency during human-machine interactions: A behavioural study using a full humanoid robot

Although previous investigations reported a reduced sense of agency when individuals act with traditional machines, little is known about the mechanisms underpinning interactions with human-like automata. The aim of this study was twofold: (1) to investigate the effect of the machine's physical appearance on the individuals' sense of agency and (2) to explore the cognitive mechanisms underlying the individuals' sense of agency when they are engaged in a joint task. Twenty-eight participants performed a joint Simon task together with another human or an automated artificial system as a co-agent. The physical appearance of the automated artificial system was manipulated so that participants could cooperate either with a servomotor or a full humanoid robot during the joint task. Both participants' response times and temporal estimations of action-output delays (i.e., an implicit measure of agency) were collected. Results showed that participants' sense of agency for self- and other-generated actions sharply declined during interactions with the servomotor compared with the human-human interactions. Interestingly, participants' sense of agency for self- and other-generated actions was reinforced when participants interacted with the humanoid robot compared with the servomotor. These results are discussed further.

Artificial scaffolding: Augmenting social cognition by means of robot technology

The concept of scaffolding refers to the support that the environment provides in the acquisition and consolidation of new abilities. Technological advancements allow for support in the acquisition of cognitive capabilities, such as second language acquisition using simple smartphone applications There is, however, one domain of cognition that has been scarcely addressed in the context of technologically assisted scaffolding: social cognition. We explored the possibility of supporting the acquisition of social competencies of a group of children with autism spectrum disorder engaged in a rehabilitation program (age = 5.8 ± 1.14, 10 females, 33 males) by designing two robot-assisted training protocols tailored to Theory of Mind competencies. One protocol was performed with a humanoid robot and the other (control) with a non-anthropomorphic robot. We analyzed changes in NEPSY-II scores before and after the training using mixed effects models. Our results showed that activities with the humanoid significantly improved NEPSY-II scores on the ToM scale. We claim that the motor repertoire of humanoids makes them ideal platforms for artificial scaffolding of social skills in individuals with autism, as they can evoke similar social mechanisms to those elicited in human-human interaction, without providing the same social pressure that another human might exert.

Towards Living Machines: current and future trends of tactile sensing, grasping, and social robotics

The development of future technologies can be highly influenced by our deeper understanding of the principles that underlie living organisms. The Living Machines conference aims at presenting (among others) the interdisciplinary work of behaving systems based on such principles. Celebrating the 10 years of the conference, we present the progress and future challenges of some of the key themes presented in the robotics workshop of the Living Machines conference. More specifically, in this perspective paper, we focus on the advances in the field of biomimetics and robotics for the creation of artificial systems that can robustly interact with their environment, ranging from tactile sensing, grasping, and manipulation to the creation of psychologically plausible agents.

Attitudes About Artificially Intelligent Interactive Voice Response Systems Using Amazon Alexa in Cardiovascular Clinics: Insights from the VOICE-COVID-19 Study

The acceptability of artificially intelligent interactive voice response (AI-IVR) systems in cardiovascular research settings is unclear. As a result, we evaluated peoples' attitudes regarding the Amazon Echo Show 8 device when used for electronic data capture in cardiovascular clinics. Participants were recruited following the Voice-Based Screening for SARS-CoV-2 Exposure in Cardiovascular clinics study. Overall, 215 people enrolled and underwent screening (mean age 46.1; 55% females) in the VOICE-COVID study and 58 people consented to participate in a post-screening survey. Following thematic analysis, four key themes affecting AI-IVR acceptability were identified. These were difficulties with communication (44.8%), limitations with available interaction modalities (41.4%), barriers with the development of therapeutic relationships (25.9%), and concerns with universality and accessibility (8.6%). While there are potential concerns with the use of AI-IVR technologies, these systems appeared to be well accepted in cardiovascular clinics. Increased development of these technologies could significantly improve healthcare access and efficiency.

Learning from humans to build social cognition among robots

Self-organized groups of robots have generally coordinated their behaviors using quite simple social interactions. Although simple interactions are sufficient for some group behaviors, future research needs to investigate more elaborate forms of coordination, such as social cognition, to progress towards real deployments. In this perspective, we define social cognition among robots as the combination of social inference, social learning, social influence, and knowledge transfer, and propose that these abilities can be established in robots by building underlying mechanisms based on behaviors observed in humans. We review key social processes observed in humans that could inspire valuable capabilities in robots and propose that relevant insights from human social cognition can be obtained by studying human-controlled avatars in virtual environments that have the correct balance of embodiment and constraints. Such environments need to allow participants to engage in embodied social behaviors, for instance through situatedness and bodily involvement, but, at the same time, need to artificially constrain humans to the operational conditions of robots, for instance in terms of perception and communication. We illustrate our proposed experimental method with example setups in a multi-user virtual environment.

Can Robots have Personal Identity?

This article attempts to answer the question of whether robots can have personal identity. In recent years, and due to the numerous and rapid technological advances, the discussion around the ethical implications of Artificial Intelligence, Artificial Agents or simply Robots, has gained great importance. However, this reflection has almost always focused on problems such as the moral status of these robots, their rights, their capabilities or the qualities that these robots should have to support such status or rights. In this paper I want to address a question that has been much less analyzed but which I consider crucial to this discussion on robot ethics: the possibility, or not, that robots have or will one day have personal identity. The importance of this question has to do with the role we normally assign to personal identity as central to morality. After posing the problem and exposing this relationship between identity and morality, I will engage in a discussion with the recent literature on personal identity by showing in what sense one could speak of personal identity in beings such as robots. This is followed by a discussion of some key texts in robot ethics that have touched on this problem, finally addressing some implications and possible objections. I finally give the tentative answer that robots could potentially have personal identity, given other cases and what we empirically know about robots and their foreseeable future.

Is biorobotics science? Some theoretical reflections

In this paper, we ask one fairly simple question: to what extent can biorobotics be sensibly qualified as science? The answer clearly depends on what 'science' means and whether what is actually done in biorobotics corresponds to this meaning. To respond to this question, we will deploy the distinction between science and so-called technoscience, and isolate different kinds of objects of inquiry in biorobotics research. Capitalising on the distinction between 'proximal' and 'distal' biorobotic hypotheses, we will argue that technoscientific biorobotic studies address proximal hypotheses, whilst scientific biorobotic studies address distal hypotheses. As a result, we argue that bioroboticians can be both considered as scientists and technoscientists and that this is one of the main payoffs of biorobotics. Indeed, technoscientists play an extremely important role in 21st-century culture and in the current critical production of knowledge. Today's world is increasingly technological, or rather, it is a bio-hybrid system in which the biological and the technological are mixed. Therefore, studying the behaviour of robotic systems and the phenomena of animal-robot interaction means analysing, understanding, and shaping our world. Indeed, in the conclusion of the paper, we broadly reflect on the philosophical and disciplinary payoff of seeing biorobotics as a science and/or technoscience for the increasingly bio-hybrid and technical world of the 21st century.

Domain-specific and domain-general neural network engagement during human-robot interactions

To what extent do domain-general and domain-specific neural network engagement generalize across interactions with human and artificial agents? In this exploratory study, we analysed a publicly available functional MRI (fMRI) data set (n = 22) to probe the similarities and dissimilarities in neural architecture while participants conversed with another person or a robot. Incorporating trial-by-trial dynamics of the interactions, listening and speaking, we used whole-brain, region-of-interest and functional connectivity analyses to test response profiles within and across social or non-social, domain-specific and domain-general networks, that is, the person perception, theory-of-mind, object-specific, language and multiple-demand networks. Listening to a robot compared to a human resulted in higher activation in the language network, especially in areas associated with listening comprehension, and in the person perception network. No differences in activity of the theory-of-mind network were found. Results from the functional connectivity analysis showed no difference between interactions with a human or robot in within- and between-network connectivity. Together, these results suggest that although largely similar regions are activated when speaking to a human and to a robot, activity profiles during listening point to a dissociation at a lower level or perceptual level, but not higher order cognitive level.

Age-Related Differences in the Perception of Robotic Referential Gaze in Human-Robot Interaction

There is an increased interest in using social robots to assist older adults during their daily life activities. As social robots are designed to interact with older users, it becomes relevant to study these interactions under the lens of social cognition. Gaze following, the social ability to infer where other people are looking at, deteriorates with older age. Therefore, the referential gaze from robots might not be an effective social cue to indicate spatial locations to older users. In this study, we explored the performance of older adults, middle-aged adults, and younger controls in a task assisted by the referential gaze of a Pepper robot. We examined age-related differences in task performance, and in self-reported social perception of the robot. Our main findings show that referential gaze from a robot benefited task performance, although the magnitude of this facilitation was lower for older participants. Moreover, perceived anthropomorphism of the robot varied less as a result of its referential gaze in older adults. This research supports that social robots, even if limited in their gazing capabilities, can be effectively perceived as social entities. Additionally, this research suggests that robotic social cues, usually validated with young participants, might be less optimal signs for older adults.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12369-022-00926-6.

Precuneus brain response changes differently during human-robot and human-human dyadic social interaction

Human–human interactions (HHI) and human–robot interactions (HRI) are compared to identify differences between cognitive processes reflecting bonding in social interactions with natural and artificial agents. We capitalize on a unique corpus of neuroimaging data (fMRI) recorded while participants freely discussed with another human or a conversational robotic head, in order to study a crucial parameter of human social cognition, namely that social interactions are adaptive bidirectional processes that evolve over time. We used linear statistics to identify regions of the brain where activity changes differently when participants carry out twelve one-minute conversations, alternating between a human and a robotic interlocutor. Results show that activity in the posterior cingulate cortex, a key region associated with social cognition, increases over time in HHI but not in HRI. These results are interpreted as reflecting a process of strengthening social bonding during repeated exchanges when the interacting agent is a human, but not a robot.

Tracking human interactions with a commercially-available robot over multiple days

Background: As research examining human-robot interaction moves from the laboratory to the real world, studies seeking to examine how people interact with robots face the question of which robotic platform to employ to collect data in situ. To facilitate the study of a broad range of individuals, from children to clinical populations, across diverse environments, from homes to schools, a robust, reproducible, low-cost and easy-to-use robotic platform is needed. Methods: We describe how a commercially available off-the-shelf robot, Cozmo, can be used to study embodied human-robot interactions in a wide variety of settings, including the user's home. We describe the steps required to use this affordable and flexible platform for longitudinal human-robot interaction studies. First, we outline the technical specifications and requirements of this platform and accessories. We then show how log files containing detailed data on the human-robot interaction can be collected and extracted. Finally, we detail the types of information that can be retrieved from these data. Results: We present findings from a validation that mapped the behavioural repertoire of the Cozmo robot and introduce an accompanying interactive emotion classification tool to use with this robot. This tool combined with the data extracted from the log files can provide the necessary details to understand the psychological consequences of long-term interactions. Conclusions: This low-cost robotic platform has the potential to provide the field with a variety of valuable new possibilities to study the social cognitive processes underlying human-robot interactions within and beyond the research laboratory, which are user-driven and unconstrained in both time and place.

Personalizing Care Through Robotic Assistance and Clinical Supervision

By 2030, the World Health Organization (WHO) foresees a worldwide workforce shortfall of healthcare professionals, with dramatic consequences for patients, economies, and communities. Research in assistive robotics has experienced an increasing attention during the last decade demonstrating its utility in the realization of intelligent robotic solutions for healthcare and social assistance, also to compensate for such workforce shortages. Nevertheless, a challenge for effective assistive robots is dealing with a high variety of situations and contextualizing their interactions according to living contexts and habits (or preferences) of assisted people. This study presents a novel cognitive system for assistive robots that rely on artificial intelligence (AI) representation and reasoning features/services to support decision-making processes of healthcare assistants. We proposed an original integration of AI-based features, that is, knowledge representation and reasoning and automated planning to 1) define a human-in-the-loop continuous assistance procedure that helps clinicians in evaluating and managing patients and; 2) to dynamically adapt robot behaviors to the specific needs and interaction abilities of patients. The system is deployed in a realistic assistive scenario to demonstrate its feasibility to support a clinician taking care of several patients with different conditions and needs.

What the study of spinal cord injured patients can tell us about the significance of the body in cognition

Although in the last three decades philosophers, psychologists and neuroscientists have produced numerous studies on human cognition, the debate concerning its nature is still heated and current views on the subject are somewhat antithetical. On the one hand, there are those who adhere to a view implying 'disembodiment' which suggests that cognition is based entirely on symbolic processes. On the other hand, a family of theories referred to as the Embodied Cognition Theories (ECT) postulate that creating and maintaining cognition is linked with varying degrees of inherence to somatosensory and motor representations. Spinal cord injury induces a massive body-brain disconnection with the loss of sensory and motor bodily functions below the lesion level but without directly affecting the brain. Thus, SCI may represent an optimal model for testing the role of the body in cognition. In this review, we describe post-lesional cognitive modifications in relation to body, space and action representations and various instances of ECT. We discuss the interaction between body-grounded and symbolic processes in adulthood with relevant modifications after body-brain disconnection.

Going Beyond the "Synthetic Method": New Paradigms Cross-Fertilizing Robotics and Cognitive Neuroscience

In so-called ethorobotics and robot-supported social cognitive neurosciences, robots are used as scientific tools to study animal behavior and cognition. Building on previous epistemological analyses of biorobotics, in this article it is argued that these two research fields, widely differing from one another in the kinds of robots involved and in the research questions addressed, share a common methodology, which significantly differs from the "synthetic method" that, until recently, dominated biorobotics. The methodological novelty of this strategy, the research opportunities that it opens, and the theoretical and technological challenges that it gives rise to, will be discussed with reference to the peculiarities of the two research fields. Some broad methodological issues related to the generalization of results concerning robot-animal interaction to theoretical conclusions on animal-animal interaction will be identified and discussed.

EEG alpha and theta signatures of socially and non-socially cued working memory in virtual reality

In this preregistered study (https://osf.io/s4rm9) we investigated the behavioural and neurological [electroencephalography; alpha (attention) and theta (effort)] effects of dynamic non-predictive social and non-social cues on working memory. In a virtual environment realistic human-avatars dynamically looked to the left or right side of a table. A moving stick served as a non-social control cue. Kitchen items were presented in the valid cued or invalid un-cued location for encoding. Behavioural findings showed a similar influence of the cues on working memory performance. Alpha power changes were equivalent for the cues during cueing and encoding, reflecting similar attentional processing. However, theta power changes revealed different patterns for the cues. Theta power increased more strongly for the non-social cue compared to the social cue during initial cueing. Furthermore, while for the non-social cue there was a significantly larger increase in theta power for valid compared to invalid conditions during encoding, this was reversed for the social cue, with a significantly larger increase in theta power for the invalid compared to valid conditions, indicating differences in the cues' effects on cognitive effort. Therefore, while social and non-social attention cues impact working memory performance in a similar fashion, the underlying neural mechanisms appear to differ.

People's dispositional cooperative tendencies towards robots are unaffected by robots' negative emotional displays in prisoner's dilemma games

The study explores the impact of robots' emotional displays on people's tendency to cooperate with a robot opponent in prisoner's dilemma games. Participants played iterated prisoner's dilemma games with a non-expressive robot (as a measure of cooperative baseline), followed by an angry, and a sad robot, in turn. Based on the Emotion as Social Information model, we expected participants with higher cooperative predispositions to cooperate less when a robot displayed anger, and cooperate more when the robot displayed sadness. Contrarily, according to this model, participants with lower cooperative predispositions should cooperate more with an angry robot and less with a sad robot. The results of 60 participants failed to support the predictions. Only the participants' cooperative predispositions significantly predicted their cooperative tendencies during gameplay. Participants who cooperated more in the baseline measure also cooperated more with the robots displaying sadness and anger. In exploratory analyses, we found that participants who accurately recognised the robots' sad and angry displays tended to cooperate less with them overall. The study highlights the impact of personal factors in human-robot cooperation, and how these factors might surpass the influence of bottom-up emotional displays by the robots in the present experimental scenario.

Tracking Anthropomorphizing Behavior in Human-Robot Interaction

Existing methodologies to describe anthropomorphism in human-robot interaction often rely either on specific one-time responses to robot behavior, such as keeping the robot's secret, or on post hoc measures, such as questionnaires. Currently, there is no method to describe the dynamics of people's behavior over the course of an interaction and in response to robot behavior. In this paper, I propose a method that allows the researcher to trace anthropomorphizing and non-anthropomorphizing responses to robots dynamically moment-by-moment over the course of human-robot interactions. I illustrate this methodology in a case study and find considerable variation between participants, but also considerable intrapersonal variation in the ways the robot is anthropomorphized. That is, people may respond to the robot as if it was another human in one moment and to its machine-like properties in the next. These findings may influence explanatory models of anthropomorphism.

Tactile Interaction with a Humanoid Robot: Effects on Physiology and Subjective Impressions

This study investigated how touching and being touched by a humanoid robot affects human physiology, impressions of the interaction, and attitudes towards humanoid robots. 21 healthy adult participants completed a 3 (touch style: touching, being touched, pointing) × 2 (body part: hand vs buttock) within-subject design using a Pepper robot. Skin conductance response (SCR) was measured during each interaction. Perceived impressions of the interaction (i.e., friendliness, comfort, arousal) were measured per questionnaire after each interaction. Participants’ demographics and their attitude towards robots were also considered. We found shorter SCR rise times in the being touched compared to the touching condition, possibly reflecting psychological alertness to the unpredictability of robot-initiated contacts. The hand condition had shorter rise times than the buttock condition. Most participants evaluated the hand condition as most friendly and comfortable and the robot-initiated interactions as most arousing. Interacting with Pepper improved attitudes towards robots. Our findings require future studies with larger samples and improved procedures. They have implications for robot design in all domains involving tactile interactions, such as caring and intimacy.

Irrelevant Robot Signals in a Categorization Task Induce Cognitive Conflict in Performance, Eye Trajectories, the N2 ERP-EEG Component, and Frontal Theta Oscillations

Understanding others' nonverbal behavior is essential for social interaction, as it allows, among others, to infer mental states. Although gaze communication, a well-established nonverbal social behavior, has shown its importance in inferring others' mental states, not much is known about the effects of irrelevant gaze signals on cognitive conflict markers during collaborative settings. Here, participants completed a categorization task where they categorized objects based on their color while observing images of a robot. On each trial, participants observed the robot iCub grasping an object from a table and offering it to them to simulate a handover. Once the robot "moved" the object forward, participants were asked to categorize the object according to its color. Before participants were allowed to respond, the robot made a lateral head/gaze shift. The gaze shifts were either congruent or incongruent with the object's color. We expected that incongruent head cues would induce more errors (Study 1), would be associated with more curvature in eye-tracking trajectories (Study 2), and induce larger amplitude in electrophysiological markers of cognitive conflict (Study 3). Results of the three studies show more oculomotor interference as measured in error rates (Study 1), larger curvatures eye-tracking trajectories (Study 2), and higher amplitudes of the N2 ERP of the EEG signals as well as higher event-related spectral perturbation amplitudes (Study 3) for incongruent trials compared with congruent trials. Our findings reveal that behavioral, ocular, and electrophysiological markers can index the influence of irrelevant signals during goal-oriented tasks.

Coordinating With a Robot Partner Affects Neural Processing Related to Action Monitoring

Robots start to play a role in our social landscape, and they are progressively becoming responsive, both physically and socially. It begs the question of how humans react to and interact with robots in a coordinated manner and what the neural underpinnings of such behavior are. This exploratory study aims to understand the differences in human-human and human-robot interactions at a behavioral level and from a neurophysiological perspective. For this purpose, we adapted a collaborative dynamical paradigm from the literature. We asked 12 participants to hold two corners of a tablet while collaboratively guiding a ball around a circular track either with another participant or a robot. In irregular intervals, the ball was perturbed outward creating an artificial error in the behavior, which required corrective measures to return to the circular track again. Concurrently, we recorded electroencephalography (EEG). In the behavioral data, we found an increased velocity and positional error of the ball from the track in the human-human condition vs. human-robot condition. For the EEG data, we computed event-related potentials. We found a significant difference between human and robot partners driven by significant clusters at fronto-central electrodes. The amplitudes were stronger with a robot partner, suggesting a different neural processing. All in all, our exploratory study suggests that coordinating with robots affects action monitoring related processing. In the investigated paradigm, human participants treat errors during human-robot interaction differently from those made during interactions with other humans. These results can improve communication between humans and robot with the use of neural activity in real-time.

Empathy and Schadenfreude in Human-Robot Teams

Intergroup dynamics shape the ways in which we interact with other people. We feel more empathy towards ingroup members compared to outgroup members, and can even feel pleasure when an outgroup member experiences misfortune, known as schadenfreude. Here, we test the extent to which these intergroup biases emerge during interactions with robots. We measured trial-by-trial fluctuations in emotional reactivity to the outcome of a competitive reaction time game to assess both empathy and schadenfreude in arbitrary human-human and human-robot teams. Across four experiments (total n = 361), we observed a consistent empathy and schadenfreude bias driven by team membership. People felt more empathy towards ingroup members than outgroup members and more schadenfreude towards outgroup members. The existence of an intergroup bias did not depend on the nature of the agent: the same effects were observed for human-human and human-robot teams. People reported similar levels of empathy and schadenfreude towards a human and robot player. The human likeness of the robot did not consistently influence this intergroup bias. In other words, similar empathy and schadenfreude biases were observed for both humanoid and mechanoid robots. For all teams, this bias was influenced by the level of team identification; individuals who identified more with their team showed stronger intergroup empathy and schadenfreude bias. Together, we show that similar intergroup dynamics that shape our interactions with people can also shape interactions with robots. Our results highlight the importance of taking intergroup biases into account when examining social dynamics of human-robot interactions.

What modulates the Mirror Neuron System during action observation?: Multiple factors involving the action, the actor, the observer, the relationship between actor and observer, and the context

Seeing an agent perform an action typically triggers a motor simulation of that action in the observer's Mirror Neuron System (MNS). Over the past few years, it has become increasingly clear that during action observation the patterns and strengths of responses in the MNS are modulated by multiple factors. The first aim of this paper is therefore to provide the most comprehensive survey to date of these factors. To that end, 22 distinct factors are described, broken down into the following sets: six involving the action; two involving the actor; nine involving the observer; four involving the relationship between actor and observer; and one involving the context. The second aim is to consider the implications of these findings for four prominent theoretical models of the MNS: the Direct Matching Model; the Predictive Coding Model; the Value-Driven Model; and the Associative Model. These assessments suggest that although each model is supported by a wide range of findings, each one is also challenged by other findings and relatively unaffected by still others. Hence, there is now a pressing need for a richer, more inclusive model that is better able to account for all of the modulatory factors that have been identified so far.

I Am Looking for Your Mind: Pupil Dilation Predicts Individual Differences in Sensitivity to Hints of Human-Likeness in Robot Behavior

The presence of artificial agents in our everyday lives is continuously increasing. Hence, the question of how human social cognition mechanisms are activated in interactions with artificial agents, such as humanoid robots, is frequently being asked. One interesting question is whether humans perceive humanoid robots as mere artifacts (interpreting their behavior with reference to their function, thereby adopting the design stance) or as intentional agents (interpreting their behavior with reference to mental states, thereby adopting the intentional stance). Due to their humanlike appearance, humanoid robots might be capable of evoking the intentional stance. On the other hand, the knowledge that humanoid robots are only artifacts should call for adopting the design stance. Thus, observing a humanoid robot might evoke a cognitive conflict between the natural tendency of adopting the intentional stance and the knowledge about the actual nature of robots, which should elicit the design stance. In the present study, we investigated the cognitive conflict hypothesis by measuring participants’ pupil dilation during the completion of the InStance Test. Prior to each pupillary recording, participants were instructed to observe the humanoid robot iCub behaving in two different ways (either machine-like or humanlike behavior). Results showed that pupil dilation and response time patterns were predictive of individual biases in the adoption of the intentional or design stance in the IST. These results may suggest individual differences in mental effort and cognitive flexibility in reading and interpreting the behavior of an artificial agent.

Natural human postural oscillations enhance the empathic response to a facial pain expression in a virtual character

Virtual reality platforms producing interactive and highly realistic characters are being used more and more as a research tool in social and affective neuroscience to better capture both the dynamics of emotion communication and the unintentional and automatic nature of emotional processes. While idle motion (i.e., non-communicative movements) is commonly used to create behavioural realism, its use to enhance the perception of emotion expressed by a virtual character is critically lacking. This study examined the influence of naturalistic (i.e., based on human motion capture) idle motion on two aspects (the perception of other’s pain and affective reaction) of an empathic response towards pain expressed by a virtual character. In two experiments, 32 and 34 healthy young adults were presented video clips of a virtual character displaying a facial expression of pain while its body was either static (still condition) or animated with natural postural oscillations (idle condition). The participants in Experiment 1 rated the facial pain expression of the virtual human as more intense, and those in Experiment 2 reported being more touched by its pain expression in the idle condition compared to the still condition, indicating a greater empathic response towards the virtual human’s pain in the presence of natural postural oscillations. These findings are discussed in relation to the models of empathy and biological motion processing. Future investigations will help determine to what extent such naturalistic idle motion could be a key ingredient in enhancing the anthropomorphism of a virtual human and making its emotion appear more genuine.

Eye contact during joint attention with a humanoid robot modulates oscillatory brain activity

Eye contact established by a human partner has been shown to affect various cognitive processes of the receiver. However, little is known about humans' responses to eye contact established by a humanoid robot. Here, we aimed at examining humans' oscillatory brain response to eye contact with a humanoid robot. Eye contact (or lack thereof) was embedded in a gaze-cueing task and preceded the phase of gaze-related attentional orienting. In addition to examining the effect of eye contact on the recipient, we also tested its impact on gaze-cueing effects (GCEs). Results showed that participants rated eye contact as more engaging and responded with higher desynchronization of alpha-band activity in left fronto-central and central electrode clusters when the robot established eye contact with them, compared to no eye contact condition. However, eye contact did not modulate GCEs. The results are interpreted in terms of the functional roles involved in alpha central rhythms (potentially interpretable also as mu rhythm), including joint attention and engagement in social interaction.

What Makes a Robot Social? A Review of Social Robots from Science Fiction to a Home or Hospital Near You

Purpose of Review

We provide an outlook on the definitions, laboratory research, and applications of social robots, with an aim to understand what makes a robot social—in the eyes of science and the general public.

Recent Findings

Social robots demonstrate their potential when deployed within contexts appropriate to their form and functions. Some examples include companions for the elderly and cognitively impaired individuals, robots within educational settings, and as tools to support cognitive and behavioural change interventions.

Summary

Science fiction has inspired us to conceive of a future with autonomous robots helping with every aspect of our daily lives, although the robots we are familiar with through film and literature remain a vision of the distant future. While there are still miles to go before robots become a regular feature within our social spaces, rapid progress in social robotics research, aided by the social sciences, is helping to move us closer to this reality.

Mind Meets Machine: Towards a Cognitive Science of Human-Machine Interactions

As robots advance from the pages and screens of science fiction into our homes, hospitals, and schools, they are poised to take on increasingly social roles. Consequently, the need to understand the mechanisms supporting human-machine interactions is becoming increasingly pressing. We introduce a framework for studying the cognitive and brain mechanisms that support human-machine interactions, leveraging advances made in cognitive neuroscience to link different levels of description with relevant theory and methods. We highlight unique features that make this endeavour particularly challenging (and rewarding) for brain and behavioural scientists. Overall, the framework offers a way to study the cognitive science of human-machine interactions that respects the diversity of social machines, individuals' expectations and experiences, and the structure and function of multiple cognitive and brain systems.

Age-related changes in gaze behaviour during social interaction: An eye-tracking study with an embodied conversational agent

Previous research has highlighted age-related differences in social perception, in particular emotional expression processing. To date, such studies have largely focused on approaches that use static emotional stimuli that the participant has to identify passively without the possibility of any interaction. In this study, we propose an interactive virtual environment to better address age-related variations in social and emotional perception. A group of 22 young (18-30 years) and 20 older (60-80 years) adults were engaged in a face-to-face conversation with an embodied conversational agent. Participants were invited to interact naturally with the agent and to identify his facial expression. Their gaze behaviour was captured by an eye-tracking device throughout the interaction. We also explored whether the Big Five personality traits (particularly extraversion) and anxiety modulated gaze during the social interaction. Findings suggested that age-related differences in gaze behaviour were only apparent when decoding social signals (i.e., listening to a partner's question, identifying facial expressions) and not when communicating social information (i.e., when speaking). Furthermore, higher extraversion levels consistently led to a shorter amount of time gazing towards the eyes, whereas higher anxiety levels led to slight modulations of gaze only when participants were listening to questions. Face-to-face conversation with virtual agents can provide a more naturalistic framework for the assessment of online socio-emotional interaction in older adults, which is not easily observable in classical offline paradigms. This study provides novel and important insights into the specific circumstances in which older adults may experience difficulties in social interactions.

Toward ethical cognitive architectures for the development of artificial moral agents

New technologies based on artificial agents promise to change the next generation of autonomous systems and therefore our interaction with them. Systems based on artificial agents such as self-driving cars and social robots are examples of this technology that is seeking to improve the quality of people’s life. Cognitive architectures aim to create some of the most challenging artificial agents commonly known as bio-inspired cognitive agents. This type of artificial agent seeks to embody human-like intelligence in order to operate and solve problems in the real world as humans do. Moreover, some cognitive architectures such as Soar, LIDA, ACT-R, and iCub try to be fundamental architectures for the Artificial General Intelligence model of human cognition. Therefore, researchers in the machine ethics field face ethical questions related to what mechanisms an artificial agent must have for making moral decisions in order to ensure that their actions are always ethically right. This paper aims to identify some challenges that researchers need to solve in order to create ethical cognitive architectures. These cognitive architectures are characterized by the capacity to endow artificial agents with appropriate mechanisms to exhibit explicit ethical behavior. Additionally, we offer some reasons to develop ethical cognitive architectures. We hope that this study can be useful to guide future research on ethical cognitive architectures.

The understanding of congruent and incongruent referential gaze in 17-month-old infants: an eye-tracking study comparing human and robot

Several studies have shown that the human gaze, but not the robot gaze, has significant effects on infant social cognition and facilitate social engagement. The present study investigates early understanding of the referential nature of gaze by comparing-through the eye-tracking technique-infants' response to human and robot's gaze. Data were acquired on thirty-two 17-month-old infants, watching four video clips, where either a human or a humanoid robot performed an action on a target. The agent's gaze was either turned to the target (congruent) or opposite to it (incongruent). The results generally showed that, independent of the agent, the infants attended longer at the face area compared to the hand and target. Additionally, the effect of referential gaze on infants' attention to the target was greater when infants watched the human compared to the robot's action. These results suggest the presence, in infants, of two distinct levels of gaze-following mechanisms: one recognizing the other as a potential interactive partner, the second recognizing partner's agency. In this study, infants recognized the robot as a potential interactive partner, whereas ascribed agency more readily to the human, thus suggesting that the process of generalizability of gazing behaviour to non-humans is not immediate.

Re-imaging the intentional stance

The commonly used paradigm to investigate Dennet's 'intentional stance' compares neural activation when participants compete with a human versus a computer. This paradigm confounds whether the opponent is natural or artificial and whether it is intentional or an automaton. This functional magnetic resonance imaging study is, to our knowledge, the first to investigate the intentional stance by orthogonally varying perceptions of the opponents' intentionality (responding actively or passively according to a script) and embodiment (human or a computer). The mere perception of the opponent (whether human or computer) as intentional activated the mentalizing network: the temporoparietal junction (TPJ) bilaterally, right temporal pole, anterior paracingulate cortex (aPCC) and the precuneus. Interacting with humans versus computers induced activations in a more circumscribed right lateralized subnetwork within the mentalizing network, consisting of the TPJ and the aPCC, possibly reflective of the tendency to spontaneously attribute intentionality to humans. The interaction between intentionality (active versus passive) and opponent (human versus computer) recruited the left frontal pole, possibly in response to violations of the default intentional stance towards humans and computers. Employing an orthogonal design is important to adequately capture Dennett's conception of the intentional stance as a mentalizing strategy that can apply equally well to humans and other intentional agents.

Shall I Trust You? From Child-Robot Interaction to Trusting Relationships

Studying trust in the context of human-robot interaction is of great importance given the increasing relevance and presence of robotic agents in the social sphere, including educational and clinical. We investigated the acquisition, loss, and restoration of trust when preschool and school-age children played with either a human or a humanoid robot in vivo. The relationship between trust and the representation of the quality of attachment relationships, Theory of Mind, and executive function skills was also investigated. Additionally, to outline children's beliefs about the mental competencies of the robot, we further evaluated the attribution of mental states to the interactive agent. In general, no substantial differences were found in children's trust in the play partner as a function of agency (human or robot). Nevertheless, 3-year-olds showed a trend toward trusting the human more than the robot, as opposed to 7-year-olds, who displayed the reverse pattern. These findings align with results showing that, for 3- and 7-year-olds, the cognitive ability to switch was significantly associated with trust restoration in the human and the robot, respectively. Additionally, supporting previous findings, we found a dichotomy between attributions of mental states to the human and robot and children's behavior: while attributing to the robot significantly lower mental states than the human, in the Trusting Game, children behaved in a similar way when they related to the human and the robot. Altogether, the results of this study highlight that similar psychological mechanisms are at play when children are to establish a novel trustful relationship with a human and robot partner. Furthermore, the findings shed light on the interplay - during development - between children's quality of attachment relationships and the development of a Theory of Mind, which act differently on trust dynamics as a function of the children's age as well as the interactive partner's nature (human vs. robot).

A neurocognitive investigation of the impact of socializing with a robot on empathy for pain

To what extent can humans form social relationships with robots? In the present study, we combined functional neuroimaging with a robot socializing intervention to probe the flexibility of empathy, a core component of social relationships, towards robots. Twenty-six individuals underwent identical fMRI sessions before and after being issued a social robot to take home and interact with over the course of a week. While undergoing fMRI, participants observed videos of a human actor or a robot experiencing pain or pleasure in response to electrical stimulation. Repetition suppression of activity in the pain network, a collection of brain regions associated with empathy and emotional responding, was measured to test whether socializing with a social robot leads to greater overlap in neural mechanisms when observing human and robotic agents experiencing pain or pleasure. In contrast to our hypothesis, functional region-of-interest analyses revealed no change in neural overlap for agents after the socializing intervention. Similarly, no increase in activation when observing a robot experiencing pain emerged post-socializing. Whole-brain analysis showed that, before the socializing intervention, superior parietal and early visual regions are sensitive to novel agents, while after socializing, medial temporal regions show agent sensitivity. A region of the inferior parietal lobule was sensitive to novel emotions, but only during the pre-socializing scan session. Together, these findings suggest that a short socialization intervention with a social robot does not lead to discernible differences in empathy towards the robot, as measured by behavioural or brain responses. We discuss the extent to which long-term socialization with robots might shape social cognitive processes and ultimately our relationships with these machines. This article is part of the theme issue 'From social brains to social robots: applying neurocognitive insights to human-robot interaction'.

Social robotics and the modulation of social perception and bias

The field of social robotics offers an unprecedented opportunity to probe the process of impression formation and the effects of identity-based stereotypes (e.g. about gender or race) on social judgements and interactions. We present the concept of fair proxy communication-a form of robot-mediated communication that proceeds in the absence of potentially biasing identity cues-and describe how this application of social robotics may be used to illuminate implicit bias in social cognition and inform novel interventions to reduce bias. We discuss key questions and challenges for the use of robots in research on the social cognition of bias and offer some practical recommendations. We conclude by discussing boundary conditions of this new form of interaction and by raising some ethical concerns about the inclusion of social robots in psychological research and interventions. This article is part of the theme issue 'From social brains to social robots: applying neurocognitive insights to human-robot interaction'.