When witnessing the abuse of others, humans generally exhibit emotional responses. A large number of studies in Human-Robot-Interaction (HRI) have shown that humans also react with sympathy when robots are abused, but most of these insights come from controlled laboratory studies using short videos and student samples. To complement existing research with observations drawn from real-world online discussions, this paper presents a sentiment analysis of 103,413 YouTube comments on videos depicting abuse of animal-like, humanoid, and cart-shaped robots. To validate our sentiment classification, we analysed the comments using a lexicon-based tool, two fine-tuned language models, and three general-purpose state-of-the-art large language models (LLMs). The comparison yielded interesting results: LLMs generally classified science-fiction–related comments, e.g., references to dystopian TV shows, as negative, while lexicon and fine-tuned models mainly labelled them as neutral. The six models agreed on the classification of a total of 27,427 comments, which we used to explore the sentiment expressions occurring across videos featuring robots with different physical forms. Our findings provide large-scale, ecologically valid insights into how emotional responses to robot abuse are expressed and analysed in online video platforms.

Human-like nonverbal behaviors, such as gestures or facial expressions, play a crucial role in face-to-face communication, yet their impact on user interactions with virtual agents (VAs) in real-world environments remains underexplored. This study examines how co-speech gestures influence engagement and perception in interactions with a VA deployed in a museum. Using a mixed-methods approach, we conducted an A/B test comparing a gesturing VA to a nongesturing VA. Quantitative analysis of system logs revealed no significant difference in the number of conversational turns between conditions. Qualitative findings revealed that gestures may positively influence usability by making the interaction more intuitive and reducing user confusion regarding the VA’s internal states. However, perceptions of nonverbal behavior varied, with some users finding gestures engaging while others considered them uncanny. Regardless of condition, system response times, verbal communication errors, and answer quality were key drivers of user dissatisfaction. These findings highlight both the benefits and challenges of incorporating and studying nonverbal behavior in VAs and underscore the need for robust verbal communication to support meaningful interactions.

The rise of generative AI, particularly large language models (LLMs), prompts a re-examination of human-agent interaction (HAI), inviting us to imagine novel roles for agents in society. This paper explores this theme of “Interaction and Imagination” through a multi-stage exploratory study on enhancing an embodied conversational agent (ECA) for older adults’ health monitoring. First, to ground our work in user needs, we conducted a co-creation workshop with older users (N=3) who had extensive, real-world experience with a pre-existing, intent-based health ECA. The goal was to identify its core limitations (limited contextual interpretation; rigid dialog) and to let users envision the key features of an ideal successor. This revealed a strong desire for an ECA with greater interactional flexibility, contextual understanding of health data, and social engagement. Second, guided by these co-created requirements based on voiced limitations and desires, we developed an LLM-enhanced prototype featuring a hybrid rule-based and generative dialog model. Finally, to evaluate whether this successor addressed the initial limitations, we conducted an exploratory within-subjects mixed-methods comparative study (N=7) directly contrasting the new LLM-ECA against the original intent-based system. Results indicate a ‘double-edged sword’: While the LLM-ECA was perceived as more ‘organic’, it was not rated higher on overall anthropomorphism and presented significant challenges in intuitiveness and cognitive load, exacerbated by user interface (UI) limitations such as push-to-talk voice activation. This study highlights the critical tension between the imagined potential of fluid, AI-driven conversation and the practical realities of interaction for older adults. We provide early empirical evidence underscoring the necessity of hybrid systems that balance generative adaptability with guided interaction, alongside robust UI design and structured onboarding. Our work offers actionable considerations for designing LLM-ECAs that effectively bridge the gap between imaginative possibilities and impactful HAI in sensitive domains.

Using social robots and virtual agents (VAs) as interfaces for health monitoring systems for older adults offers the possibility of more engaging interactions that can support long-term health and well-being. While robots are characterized by their physical presence, software-based VAs are more scalable and flexible. Few comparisons of these interfaces exist in the human-robot and humanagent interaction domains, especially in long-term and real-world studies. In this work, we examined impressions of social robots and VAs at the beginning and end of an eight-week study in which older adults interacted with these systems independently in their homes. Using a between-subjects design, participants could choose which interface to evaluate during the study. While participants perceived the social robot as somewhat more likable, the VA was perceived as more intelligent. Our work provides a basis for further studies investigating factors most relevant for engaging interactions with social interfaces for long-term health monitoring.

Stroke-induced aphasia, an acquired speech impairment, poses significant challenges to individuals’ communication abilities. Our robot-assisted speech training app aims to facilitate home-based, self-administered training for individuals with aphasia to complement their speech therapy. We evaluated our app in a single-session study with four individuals with aphasia in a rehabilitation setting. Each participant had a different severity grade of the condition, including one case of global aphasia. Our findings suggest an overall positive user experience, with indications that the training facilitated by the robot is suitable for individuals across all severity grades of aphasia, though exercise customization is crucial. Participants, even those with limited technical experience, adapted to using the system by themselves quickly. However, usability issues specific to this diverse target group were noted, such as the length of explanations and the abundant use of robot gestures, which will be addressed in future iterations.

Social robots and virtual agents can provide a low-threshold social interface for the independent self-health monitoring of older adults, thereby supporting their self-care and empowering them to take charge of their own health. In this work, we present a study on the interface preference of 35 older adults for a health monitoring system which they will be evaluating for eight weeks in their own homes. We found that participants who prefer the social robot have a higher affinity for technology interaction (ATI) score and are more likely to use assistance systems in their daily lives, while participants with a slightly lower ATI score prefer the virtual agent. Participants cited more personality and an interest in robots as reasons for the robot preference, and space and flexibility as reasons to prefer the virtual agent. These results underscore the importance of a personalized introduction of social technologies for health monitoring into the daily lives of older adults.

Aphasia is a common symptom of stroke. Patients affected may experience difficulties in all aspects of language. To complement traditional logopedic therapy, we developed a speech training application for a social robot, allowing affected individuals to perform additional training independently. In an early evaluation, a representative of each main stakeholder group we identified - namely persons with aphasia, care staff, and speech therapists - evaluated our application in terms of perceived usefulness, ease of use, and overall user experience. The robot guided the participants through the training session autonomously and most exercises were completed without help, thus proving the feasibility of our concept. The participants rated the application overall as positive and we achieved promising results in terms of attitude towards and intention to use the system. Furthermore, the social component of the training with the robot was very well received among the participants. In the future, the training content will be revised with the help of a linguist, to adequately support the training needs of persons with aphasia.