Science AI Journal
Psychology · Research topic

Open research questions in Social Robot Interaction and HRI

55 unresolved questions extracted from the limitations and future-work sections of 1,051 Social Robot Interaction and HRI papers in our library. Each links back to the study that raised it.

What the literature leaves open

  • https://doi.org/10.1016/j.compedu.2018.07.020 of Crompton, H., Gregory, K., & Burke, D. (2018). Humanoid robots supporting children’s learning in an early childhood setting. British Journal of Educational Technology, 49(5), 911-927. https://doi.org/10.1111/bjet.12654 Dautenhahn, K., & Billard, A. G. (1999). Bringing up robots or—the psychology of socially intelligent robots: From theory to 366-367).

    Social robots disruptive artificial intelligence educational technology: Pre-service teachers’ perceptions in China, Russia and Slovenia · 2026 · DOI

  • Our paper explored the academic debate on queering sex robots. First, our paper pro- vided philosophical foundations for queering sex robots. To do so, we have explored Queer Lacanian Psychoanalysis and New Materialism in tandem. These theories offer potential for a fluid conception of human sexual identity, emphasize our close relationship with non-humans (e.g., technologies), and can thus inform our “cyborg sexuality” —, that is, our entanglement with technology, in sexuality — while also criticizing purity ideals. Second, beyond this theoretical contribution, we aimed to reflect on propositions to further imagine queering sex robots based on the theoretical part. This includes non- humanoid and fluid designs of sex robots, designing sex robots for friction, focusing on whole-body experiences and alternative forms of sexual interaction, alternative materials, and, on a more general level, critically addressing dataand development processes. However, this also implies a fundamental shift in what has so far been understood as a sex robot, raising questions about the boundary between sex robots and sex toys. Nevertheless, even though we believe that sex robots could be developed in a way that empowers female and queer individuals and challenges heteronormative views regarding sexuality, it remains important to be critical of sex robots, especially when it comes to potential abuse or data security in these interactions. Queering Sex Robots Beyond Diversifying Design? Insights from Queer…1 3 47 Page 22 of 26 The social exclusion and political prosecution that queer individuals can face makes this need for data protection of utmost importance. Furthermore, a focus on designing sex robots differently should not distract attention from emancipatory work on changing the broader social structure in which sex robots are developed, marketed and become products of desire, to ensure that queer sex robots do not serve as techno-solutions to easily brush over the structural social inequalities in which they are developed. This paper should still be understood as a starting point for queering sex robots, which needs to be further explored through interdisciplinary collaborative efforts of philosophers, ethicists, designers, and artists. For future research, it is also crucial to take a more explicitly intersectional lens, for example, the intersection of race and gender to understand the diverse experiences and challenges faced by individuals in relation to sexual technologies. Acknowledgments We gratefully acknowledge the valuable feedback we received following our pre- sentation at the Society for Philosophy and Technology Conference 2025 conference in Eindhoven, the Netherlands. Author Contributions MvdH (first author); AP (second author). Both authors jointly developed, wrote, and revised the manuscript. Funding AP is affiliated with the research programme Ethics of Socially Disruptive Technologies (ESDiT). This research programme is funded through the Gravitation programme of the Dutch Ministry of Education, Culture, and Science and the Netherlands Organization for Scientific Research (NWO grant number 024.004.031). Data Availability Not applicable.

    Queering Sex Robots Beyond Diversifying Design? Insights from Queer Lacanian Psychoanalysis and New Materialism · 2026 · DOI

  • By providing a clear conceptual framework linking participation, group interventions, and the NAO robot, the study enhances the existing literature with a synthetic proposal that has not been systematically addressed in the past.

    Enhancing the Engagement of Children with Autism Spectrum Disorder through Group-Based Interventions Utilizing the Social Robot NAO: A Synthetic Conceptual Analysis · 2026 · DOI

  • The confluence of cobot technology with AI, digital twins, and IIoT infrastructure suggests that future research should focus on the integrated human-robot-digital system rather than individual cobot deployments. The variety of outcome measures across the included studies, due to the lack of standardized cobot performance benchmarking methodologies, limits the precision and comparability of quantitative results. The Technology Acceptance Model, while useful for predicting individual operator acceptance, is insufficient as a sole explanatory framework for cobot integration outcomes.

    The Impact of Collaborative Robots in Manufacturing · 2026 · DOI

  • As the psychological effects of technology on remoteness (and its consequent behavior) are thus far understudied, we had participants (N = 205) complete an ostensible insect-killing task in which participants were either close (in the same room) or remote (via webcam) from their targets.

    Remoteness decreases negative feelings about killing · 2026 · DOI

  • Future work should investigate customizable action vocabularies or hybrid approaches that combine predefined actions with natural language extensions. Our tasks involved only a few goals with relatively simple sequences, so future work should investigate how Distill scales to complex multi-step tasks with many goals, intricate dependencies, and conditional logic. Similarly, we evaluated Distill exclusively in simulated delivery tasks, meaning that the ability of the approach to generalize to other domains remains to be seen. An important open question is for whom and under what conditions this cost of Distill is justified.

    Distill: Uncovering the True Intent behind Human-Robot Communication · 2026 · DOI

  • Research on the self model for embodied AI is still at an early stage. Existing efforts focus on individual components and remain fragmented rather than orga- nized into an integrated, testable framework. Against this backdrop, the hierarchy and the prototype sys- tem presented in this paper are an attempt oriented toward feasibility and standardization. Moreover, the method of self modeling is not a single fixed pathway. Depending on the platform and application scenario, future systems may adopt various forms that bridge symbolic and subsymbolic representations.

    Self Model for Embodied Artificial Intelligence · 2026 · DOI

  • The paper identifies VR/AR integration as a potential enhancement for immersion in telepresence systems but provides no baseline metrics, implementation approach, or assessment framework; the specific technical integration pathways and resulting pedagogical impact remain unspecified.

    Power-Optimized AI-Enhanced Telepresence Robots: A Validated Multi-Modal Framework for Sustainable Remote Learning in Higher Education · 2026 · DOI

  • Battery endurance is currently unspecified in tested configurations; expansion to 8-10 hour operational periods using 6000 mAh high-capacity cells requires hardware redesign and power-consumption validation to confirm the 35% energy reduction advantage is maintained.

    Power-Optimized AI-Enhanced Telepresence Robots: A Validated Multi-Modal Framework for Sustainable Remote Learning in Higher Education · 2026 · DOI

  • Volunteer recruitment of educators and students introduces self-selection bias toward technology-inclined users; the system's effectiveness with stratified populations varying in digital literacy levels and technology receptiveness has not been evaluated through randomized or stratified sampling designs.

    Power-Optimized AI-Enhanced Telepresence Robots: A Validated Multi-Modal Framework for Sustainable Remote Learning in Higher Education · 2026 · DOI

  • The 315-participant sample provided cross-sectional engagement and usability data but lacks longitudinal assessment; sustained participation rates, learning outcome retention, and adaptive AI effectiveness over extended deployment periods (semester-long or academic-year-long studies) are unmeasured.

    Power-Optimized AI-Enhanced Telepresence Robots: A Validated Multi-Modal Framework for Sustainable Remote Learning in Higher Education · 2026 · DOI

  • The telepresence robot system was tested exclusively at 50 Mbps Wi-Fi connectivity; performance degradation at lower bandwidth speeds (e.g., 5 Mbps typical in rural classroom locations) remains uninvestigated, potentially overestimating NLP accuracy and latency metrics in resource-constrained educational environments.

    Power-Optimized AI-Enhanced Telepresence Robots: A Validated Multi-Modal Framework for Sustainable Remote Learning in Higher Education · 2026 · DOI

  • this principle extends to other domains in which generative agents can positively influence human decision-making, such as motivating students to learn. At the same time, restricting strategically access exploitable knowledge to sensitive or impact of misaligned agents. Taken may reduce the harmful together, these insights may inform future developments of GSAs that are both effective and aligned with human autonomy and wellbeing.

    Never say never: exploring the effects of knowledge availability on agent persuasiveness in controlled physiotherapy motivation dialogues · 2026 · DOI

  • This paper presents a novel methodological approach to investigate the persuasive behavior of GSAs, combining qualitative and quantitative means. Because GSAs imitate human behavior based on vast amounts of human-generated data, their behavioral patterns may be investigated from a psychological perspective as well. As such, the current research switched the focus from predefined agent behaviors to observing autonomously generated persuasive strategies by using fixed human messages as input. While this method allowed us to compare different knowledge configurations in a safe and controlled manner, it also raises questions about the validity of the interactions for real-life physiotherapy settings, where human messages have higher variability. However, the intent behind the scenarios was not to create universal settings for physiotherapy motivation— as this would have been impossible due to nearly unlimited possibilities in open-ended interactions. Rather, the goal was to use a specific example with high relevance as a means to investigate the agent’s autonomous decision-making. Hence, in order to observe which persuasive strategies and behaviors the GSA would choose on its own, we framed the motivation task relatively broadly without specifying behavioral rules. In more applied settings, future research should focus more on the effects of specific behavioral rules (e.g., using at most three persuasion attempts). Such studies should also equip the agent with a broader range of non-verbal expressive capabilities than the purely text- based approach implemented in the current work. Nevertheless, the present work lays the foundation for future research in such directions.

    Never say never: exploring the effects of knowledge availability on agent persuasiveness in controlled physiotherapy motivation dialogues · 2026 · DOI

  • Users reported difficulty interrupting IVAs during long responses in immersive XR environments. Future work must investigate and implement effective interruption mechanisms that preserve conversational naturalness while maintaining social presence in extended human-IVA interactions.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • Participants with neurodivergence (e.g., Asperger's syndrome) reported difficulty interpreting IVA facial expressions, indicating need for customizable emotional expressiveness. Future development should implement adjustable facial emotion clarity, exaggeration levels, and voice-synchronized expression in intelligent virtual agents for neurodivergent users.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • IVA gaze behavior and eye contact produced conflicting user responses based on cultural background (e.g., Japanese vs. USA participants perceiving prolonged eye contact differently). Future work must develop culturally-adaptive IVA designs with user-adjustable gaze behavior and agent proximity parameters tailored to individual and cultural preferences.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • User diversity in evaluations was limited, with participants predominantly being field experts, limiting generalizability of findings to non-expert populations. Future studies must include diverse user groups with varying backgrounds, expertise levels, and cultural contexts interacting with anthropomorphic virtual agents.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • Pilot evaluation of the toolkit included only a small number of developer participants due to early deployment stage, yielding only descriptive exploratory findings. Future evaluations must employ hypothesis-based testing in controlled environments to rigorously assess both the toolkit and IVAs generated with it.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • Extended in-the-wild human-AI interactions in immersive XR environments have not been adequately evaluated for maintaining social presence over longer periods. Specific design aspects requiring investigation include natural turn-taking mechanisms, user support for topic generation when uncertain, and contextual adaptation across the reality-virtuality spectrum in IVA design.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • The current Gemini streaming API integration does not support modular customization, preventing developers from combining third-party voice services (e.g., ElevenLabs) with alternative LLM/VLM models from different cloud providers. Future work must develop a modular architecture for intelligent virtual agents that enables flexible composition of STT, TTS, and LLM components.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • STT models currently struggle with users having strong accents, causing false content generation in IVA interactions. Future development must integrate accent-robust STT models that can handle diverse phonetic variations to improve accessibility and reduce user frustration in anthropomorphic AI systems.

    Anthropomorphic AI: a toolkit for authoring and interacting with intelligent virtual agents for extended reality · 2026 · DOI

  • The thesis will further argue that Malle’s approach is insufficient due to three reasons: his function argument is very simplifying and therefore troubling; each component of his theory is inconsistent and, finally, closely connected to our common understanding of personhood, which raises new philosophical questions surrounding the basic issue of if and/or when machines can be considered people.

    Robot Morality: Bertram F. Malle’s Concept of Moral Competence · 2018 · DOI

  • Although we refer to athletes who have outstanding sporting performances as machines, for example, in cycling people say ‘the cyclist looked like a machine with wheels,’ the potential participation of such AI in sport has not been addressed.

    Will robots ever play sports? · 2016 · DOI

  • However, prior research has yielded mixed findings regarding how loneliness shapes consumer responses to anthropomorphic AI service robots.

    Social Versus Emotional Loneliness: Diverging Consumer Responses to Anthropomorphic AI Service Robots · 2026 · DOI

Most-cited papers in Social Robot Interaction and HRI

Most recent work

Find a gap in your own Social Robot Interaction and HRI sub-topic

This page shows what the Social Robot Interaction and HRI literature already flags as unresolved. To narrow it to your specific question, run the guided finder — it searches the gap library on demand and checks candidates against 250M+ OpenAlex works.

Open the Research Gap Finder →

Related topics in Psychology

LGBTQ Health, Identity, and Policy
750 open questions
Child and Adolescent Psychosocial and Emotional Development
645 open questions
Reading and Literacy Development
599 open questions
Posttraumatic Stress Disorder Research
543 open questions
Suicide and Self-Harm Studies
482 open questions
Child Abuse and Trauma
462 open questions
Attachment and Relationship Dynamics
453 open questions
Family and Disability Support Research
453 open questions
55 open questions have been extracted from the limitations and future-work passages of 1,051 Social Robot Interaction and HRI papers in our library. Each one below links back to the study that raised it, so you can read the original claim in context.

Tools for your next paper

Gaps & Literature SearchFind open questions in any field — 120,000+ open gaps mined from our own library.Pre-CheckIs your paper ready? A Tier 1–5 acceptance probability in seconds.Journal FinderA ranked shortlist of target journals from a 17,500-venue index.Duplicate Publication CheckerCheck prior-publication & salami-slicing risk across 8 sources.Citation GeneratorFreeA DOI or title → APA, MLA, Vancouver, Chicago, BibTeX, RIS. No signup.Graphical Abstract MakerFreeTurn your findings into a colorblind-safe graphical abstract.

Compare the category — Honest roundups of the AI research tools, ours listed alongside the alternatives.

Command palette

Jump anywhere, run any action.