A key challenge in social robotics is identifying the design features and social mechanisms that sustain long-term engagement with robots. Although mounting evidence suggests that end-user customization is a vital aspect of robot ownership “in the wild”, empirical research on this phenomenon and its psychological outcomes remains sparse. In this mixed methods study, we surveyed 113 robot owners and conducted semi-structured interviews with 13 more, providing a holistic perspective on customization practices among long-term users. Our findings show that customization is highly prevalent among robot owners, with quantitative results indicating that customization indirectly predicts robot attachment through self-extension and psychological ownership. Our interviews furthermore reveal a vibrant culture of customization in online and offline robotics spaces, which is sustained by strong community networks. Together, these results underscore the central role of customization in fostering enduring engagement with companion robots. Through customization, users imbue their robots with personally significant identities, form deeper attachments to them, and reinforce their own individuality as robot owners. Customization also embeds owners within a broader community of robot enthusiasts, promoting social connections, creative practice, and sustained use. Given its prevalence among robot owners, we conclude with recommendations for how robot designers and researchers can leverage customization to set the stage for long-term and personally meaningful human–robot bonds.

Self-disclosure and the social sharing of emotions facilitate social relationships and can positively affect people's well-being. Nevertheless, individuals might refrain from engaging in these interpersonal communication behaviours with other people, due to socio-emotional barriers, such as shame and stigma. Social robots, free from these human-centric judgements, could encourage openness and overcome these barriers. Accordingly, this paper reviews the role of self-disclosure and social sharing of emotion in human-robot interactions (HRIs), particularly its implications for emotional well-being and the dynamics of social relationship building between humans and robots. We investigate the transition of self-disclosure dynamics from traditional human-to-human interactions to HRI, revealing the potential of social robots to bridge socio-emotional barriers and provide unique forms of emotional support. This review not only highlights the therapeutic potential of social robots but also raises critical ethical considerations and potential drawbacks of these interactions, emphasising the importance of a balanced approach to integrating robots into emotional support roles. The review underscores a complex but promising frontier at the intersection of technology and emotional well-being, advocating for careful consideration of ethical standards and the intrinsic human need for connection as we advance in the development and application of social robots.

Autonomous drones are expected to become prevalent in populated environments such as warehouses, factories, and homes, where individuals and drones will coexist while independently performing tasks. However, the dynamics of this shared autonomy, particularly the spatial behaviors (proxemics) that arise in these settings, remain underexplored in the Human–Drone Interaction (HDI) field. Understanding how task-driven behaviors influence navigation around drones is critical for ensuring their seamless integration into such environments. This study investigates how individuals’ prioritization of task completion interacts with proxemic behaviors, potentially overriding defensive mechanisms like maintaining safety distances. We conducted a study with 26 participants in a fully immersive virtual environment where both participants and a drone were tasked with moving objects. Using a 2 × 2 within-subject design, we examined proxemics in relaxed and competitive scenarios, with the drone carrying either normal or explosive boxes (triggering explosions on contact). Results revealed a significant interaction between scenario type and the drone’s danger level: in relaxed scenarios, participants deviated more from the shortest path when the drone carried explosive boxes. However, stronger task-oriented behavior correlated with closer approaches to the drone, regardless of danger. These findings highlight the dominance of goal-oriented behavior in shaping Human–Drone Proxemics and its implications for drone deployment in high-performance environments.

This study explores the impact of socio-economic and cultural factors on handwashing behaviour in schools through a robot-assisted intervention. We evaluated hand hygiene compliance across three schools, one in India and two in the UK, representing different cultural and socio-economic backgrounds, using the “WallBo” social robot to guide students (n=77) through proper handwashing techniques. The results revealed that students from lower socio-economic backgrounds demonstrated greater initial learning gains, particularly in the Indian school where the novelty of the robot contributed to heightened engagement. However, these gains were not sustained post-intervention, underscoring the importance of continuous reinforcement strategies in under- resourced settings. In contrast, students from higher-income schools demonstrated more consistent retention, likely due to stronger baseline knowledge and greater familiarity with technology. These findings underscore the importance of contextualising technology-driven interventions using social robots within socio-cultural frameworks to maximise long-term impact. Additionally, this study highlights the potential of social robots as an effective educational tool in diverse school environments, provided that long-term reinforcement strategies are put in place.

Subjective self-disclosure is an important feature of human social interaction. While much has been done in the social and behavioural literature to characterise the features and consequences of subjective self-disclosure, little work has been done thus far to develop computational systems that are able to accurately model it. Even less work has been done that attempts to model specifically how human interactants self-disclose with robotic partners. It is becoming more pressing as we require social robots to work in conjunction with and establish relationships with humans in various social settings. In this paper, our aim is to develop a custom multimodal attention network based on models from the emotion recognition literature, training this model on a large self-collected self-disclosure video corpus, and constructing a new loss function, the scale preserving cross entropy loss, that improves upon both classification and regression versions of this problem. Our results show that the best performing model, trained with our novel loss function, achieves an F1 score of 0.83, an improvement of 0.48 from the best baseline model. This result makes significant headway in the aim of allowing social robots to pick up on an interaction partner’s self-disclosures, an ability that will be essential in social robots with social cognition.

As social robots and other artificial agents become more conversationally capable, it is important to understand whether the content and meaning of self-disclosure towards these agents changes depending on the agent’s embodiment. In this study, we analysed conversational data from three controlled experiments in which participants self-disclosed to a human, a humanoid social robot, and a disembodied conversational agent. Using sentence embeddings and clustering, we identified themes in participants’ disclosures, which were then labelled and explained by a large language model. We subsequently assessed whether these themes and the underlying semantic structure of the disclosures varied by agent embodiment. Our findings reveal strong consistency: thematic distributions did not significantly differ across embodiments, and semantic similarity analyses showed that disclosures were expressed in highly comparable ways. These results suggest that while embodiment may influence human behaviour in human–robot and human–agent interactions, people tend to maintain a consistent thematic focus and semantic structure in their disclosures, whether speaking to humans or artificial interlocutors.

The promise of social robot applications for children’s education has attracted growing enthusiasm over the past decade, with the potential to augment and support diverse learning outcomes. However, the adoption of education robots and their expected benefits for children are yet to be realised, due to complexity, cost, and variability between robots. Soft robots offer a possible solution. However, a concern is that these robots may be seen as less competent, decreasing their adoption and utility in learning environments. In this preregistered, mixed-methods study, we investigated teachers’ (n = 120) perception of 12 hard and soft social robots along different dimensions, learning tasks, roles, and contexts. Teachers perceived hard robots as more competent, human-like, and familiar than soft robots. Soft robots were perceived as more physically/visually warm. Hard robots were also more likely to be perceived as suitable for "technical tasks" and adopting a teacher/tutor role for supporting the learning of adults or groups. Soft robots were more likely to be evaluated as suitable for use with younger learners in individual learning contexts and playing the role of a co-learner/novice. This study provides a detailed account of how soft and hard robot features influence teachers’ perceptions of robot suitability for education applications. The findings directly inform how to optimise the design and situation of social robots to maximize adoption, effectiveness, and accessibility across diverse learners and learning contexts. By highlighting the nuanced trade-offs between competence and warmth, this research challenges theoretical assumptions that complex hard robots are universally superior in educational settings.

The social robotics community is increasingly embracing human-centered techniques to design robots that align with users’ needs, preferences, and lived experiences. However, given the known challenges of incorporating laypeople into a design process, little empirical work has tested whether these techniques generate robotic concepts that are accepted and understood by a wider audience. In this mixed-methods online study, we examined how laypeople perceived and evaluated robots that were created through human-centered design. Fifty-two participants assessed a set of laypeople-created healthcare, education, entertainment, and telepresence robot designs according to how successfully each design signaled its intended use case. Our findings demonstrate that layperson-designed robots efficiently communicated their use context. Thus, low-level creative prototyping with end-users can be an effective way of eliciting strong initial design concepts as part of the human-centered design process, though this is influenced by the robot’s application domain. Furthermore, we showcase that simplistic robotic designs are sufficient to cue diverse affordances, highlighting the importance of matching a robot’s appearance to its intended use case. Our findings contribute to the study of human-centered design within social robotics, assessing the tools and activities end-users need to be able to meaningfully contribute to robotic design.

The potential for drones to engage with people and find applications in social contexts is often constrained by their mechanical design, leading to concerns and negative associations that significantly affect people’s acceptance of drones in their personal space. In this study, we explore how social cues can impact the spatial dynamic of Human–Drone Interaction. Within a fully immersive virtual environment, 25 participants engaged with and navigated around a drone that communicated several facial emotions on a digital display such as Joy, Sadness and Anger, and enacted distinct “gaze” behaviors (i.e., following participants or averting its gaze). Our results indicate that participants responded to the drone’s gaze in a manner akin to what is reported during human interaction: maintaining a greater distance when the drone established eye contact and instinctively getting closer when it averted its gaze. A more granular analysis revealed that participants who lacked prior familiarity with drones or possessed neutral to positive attitudes toward them demonstrated a higher sensitivity to the drone’s digital facial emotions. This finding highlights the potential for leveraging social cues to facilitate the integration of drones into various human-centric environments and tailor their design and behavior to suit specific individuals and situations.

How a robot moves is among the first things an observer notices when they encounter a robot. While considerable research has investigated the perception of robot body language, no studies yet, to our knowledge, have explored the social effects of how a robot moves through space (its gait) on people’s first impressions of a robot. To this end, here we performed two complementary experiments online (n = 98) and in-person (n = 26), with the objective of determining the extent to which a quadruped robot’s gait influences a) what animal people perceived it to be; and b) its social attributes in terms of warmth, competence, causing discomfort and zoomorphism. Results differed depending on whether participation was online or in-person, with gaits influencing participants’ perception more markedly when they encountered the robot in-person. Online, most participants saw the robot as a dog for every gait except one, while in-person participants reported more varied responses. Participants in both studies rated the more active, "bouncy" gait as warmer and less discomforting. In-person participants also consistently rated all gaits as warmer, more competent, less discomforting and more zoomorphic than did online participants. The study supports findings that in-person exposure and embodiment affect a robot’s social perception and further suggests that gait may have a limited effect as well.

To facilitate easy, seamless, and dynamic interactions between humans and social robots, it is important that the robot’s physical appearance gives clear cues to its affordances. However, little is known about what underlying concepts and assumptions shape laypeople’s perception and understanding of different robotic embodiments. To explore what robot-inexperienced users expect robots in different application domains to look like, we drew on Research through Design principles to pilot a tangible design kit with which laypeople could prototype robot designs. 27 participants with no background in robotics were asked to design robots for four different application domains. These participants were then further interrogated about their design choices in structured qualitative interviews. The resulting designs primarily ranged from mechanical to humanoid in appearance, though several animal-like entertainment robots were also created. The inductive thematic analysis of participant interviews revealed complex opinions on anthropomorphism in robotic designs and stressed social robots’ potential for customization and accessibility. Our findings provide qualitative insight into the beliefs that underlie laypeople’s understanding of robotic design in different contexts. Furthermore, we piloted a design toolkit that allows laypeople at any level of creative ability to design and critically reflect on robotic concepts and embodiments.

Self-disclosing to others can benefit emotional well-being, but socio-emotional barriers can limit people’s ability to do so. Self-disclosing towards social robots can help overcome these obstacles as robots lack judgment and can establish rapport. To further understand the influence of affective factors on people’s self-disclosure to social robots, this study examined the relationship between self-disclosure behaviour towards a social robot and people’s emotional states and their perception of the robot’s responses as comforting (i.e., being emphatic). The study included 1160 units of observation collected from 39 participants who conversed with the social robot Pepper (SoftBank Robotics) twice a week for 5 weeks (10 sessions in total), answering three personal questions in each session. Results show that perceiving the robot’s responses as more comforting was positively related to self-disclosure behaviour (in terms of disclosure duration in seconds, and disclosure length in number of words), and negative emotional states, such as lower mood, and higher feelings of loneliness and stress, were associated with higher rates of self-disclosure towards the robot. Additionally, higher rates of introversion significantly predicted higher rates of self-disclosure towards the robot. The study reveals the meaningful influence of affective states on how people behave when talking to social robots, especially when experiencing negative emotions. These findings may have implications for designing and developing social robots in therapeutic contexts.

The potential for robots as an education support tool is being rapidly realized. However, much of the existing research with education robots has involved studies which arbitrarily select robots for interventions without a foundational understanding of the features that make them best suited to serve and meet the expectations and needs of users. This study explored how children’s perceptions, expectations and experiences were shaped by aesthetic and functional features during interactions with three different, commercially-available robot ‘reading buddies’. We collected a range of quantitative and qualitative measures of subjective experience before and after children read a book with a robot of their choice. Overall, our findings indicated that social robots do indeed show strong potential to promote reading engagement in children. This was supported by robot features that signaled the perception of robots as intelligent, literate and attentive. Such features included the robot’s ability to speak and react to the story plot in a way that was both emotionally- and temporally-appropriate, so as to engage but not distract children when reading. As such, controlling the timing of robot animations during reading activities – either using human-control methods or automation – presents a key challenge in realizing the effective deployment of robots to promote reading engagement in children, particularly for those who experience reading difficulty and associated reading anxiety.

Since interactions with social robots are novel and exciting for many people, one particular concern in this specific area of human-robot interaction (HRI) is the extent to which human users will experience the interactions positively over time, when the robot’s novelty is particularly salient. In the current paper, we investigated users’ experience in long-term HRIs; how users perceive the ongoing interactions and the robot’s ability to sustain it over time. Therefore, here we examine the effect of the repeated measures (10 testing sessions) and the discussion theme (Covid-19 related vs general) on the way participants experienced the interaction quality with a social robot and perceived the robot’s communication competency over time. We found that despite individual differences between the participants, over time participants found the interactions with Pepper to be of higher quality and that Pepper’s communication skills got better. Nevertheless, our results also stressed that the discussion theme has no meaningful nor significant effect on the way people perceive Pepper and the interaction.

In this online study, we investigated how well people could recognize emotions displayed by video recordings of a Cozmo robot, and the extent to which emotion recognition is shaped by individuals' empathic traits. We also explored whether participants who report more empathic tendencies experienced more emotional contagion when watching Cozmo's emotional displays, since emotion contagion is a core aspect of empathy. We tested participants' perceptions of Cozmo's happiness, anger, sadness, surprise, and neutral displays. Across 103 participants, we report high recognition rates for most emotion categories except neutral animations. Furthermore, the mixed effects modelling revealed that an empathy subtype (the empathic concern subscale from the Interpersonal Reactivity Index) significantly impacted emotional contagion. Contrary to predictions, participants with high empathic concern subscale scores were less likely to find the robot's videos emotionally contagious. The study validates the utility of Cozmo robots to display emotional cues recognizable to human users, and further suggests that empathic traits could shape our affective interactions with robots, though perhaps in a counterintuitive way.

One limitation of social robots has been the ability of the models they operate on to infer meaningful social information about people's subjective perceptions, specifically from non-invasive behavioral cues. Accordingly, our paper aims to demonstrate how different deep learning architectures trained on data from human-robot, human-human, and human-agent interactions can help artificial agents to extract meaning, in terms of people's subjective perceptions, in speech-based interactions. Here we focus on identifying people's perceptions of their subjective self-disclosure (i.e., to what extent one perceives to be sharing personal information with an agent). We approached this problem in a data-first manner, prioritizing high quality data over complex model architectures. In this context, we aimed to examine the extent to which relatively simple deep neural networks could extract non-lexical features related to this kind of subjective self perception. We show that five standard neural network architectures and one novel architecture, which we call a Hopfield Convolutional Neural Network, are all able to extract meaningful features from speech data relating to subjective self-disclosure.

In this paper we present our preliminary results from the first trial conducted with “WallBo” a robotic buddy to improve handwashing for children in schools. The one-week trial was carried out in a Scottish school with 16 pupils, aged 6-7 in an ecologically valid setting. The 1:1 interaction with WallBo resulted in 86.25% handwashing compliance, a 33.25% improvement from the baseline handwashing technique pre-WallBo training, and an overall, ≈ 35% improvement on knowledge about hand hygiene. We also report some insights about perceptions about WallBo in this paper.