A Decade of Human-Robot Interaction Through Immersive Lenses: Reviewing Extended Reality as a Research Instrument in Social Robotics

Over the past decade, Extended Reality (XR), including Virtual, Augmented, and Mixed Reality, gained attention as a research instrument in human-robot interaction studies, but remains underexplored in empirical investigations of social robotics. To map the field, we systematically reviewed empirical studies from 2015 to 2025. Of 6,527 peer-reviewed articles, only 33 met strict inclusion criteria. We examined (1) how XR and virtual social robots are used, focusing on the software and hardware employed and the application contexts in which they are deployed, (2) data collection and analysis methods, (3) demographics of the researchers and participants, and (4) the challenges and future directions. Our findings show that social XR-HRI research is still driven by laboratory simulations, while crucial specifications - such as the hardware, software, and robots used - are often not reported. Robots typically act as passive and hardly interactive visual stimulus, while the rich biosignal (e.g., eye-tracking) and logging (e.g. motion capturing) functions of modern head-mounted displays remain largely untapped. While there are gaps in demographic reporting, the research teams and samples are predominantly tech-centric, Western, young, and male. Key limitations include hardware delays, small homogeneous samples, and short study cycles. We propose a four-phase roadmap to establish social XR-HRI as a reliable research medium, which includes (1) strengthen application contexts, (2) more robust and testable technological iterations, (3) embedding diversity in samples and research teams, and (4) the need for reporting standards, e.g., in form of a suitable taxonomy. Advancing in these directions is essential for XR to mature from a lab prototype into an ecologically valid research instrument for social robotics.

💡 Research Summary

This paper presents a systematic literature review (SLR) of how extended reality (XR)—encompassing virtual, augmented, and mixed reality—has been employed as a research instrument in social human‑robot interaction (HRI) over the past decade (2015‑2025). Starting from a pool of 6,527 peer‑reviewed articles identified through IEEE Xplore, ACM DL, Scopus, and Web of Science, the authors applied PRISMA‑guided inclusion and exclusion criteria to isolate empirical studies that (a) focus on social HRI (i.e., communication, relationship‑building, trust, cooperation) and (b) explicitly use XR as a methodological tool (simulation, evaluation, or prototyping). After duplicate removal, title/abstract screening, and full‑text assessment, only 33 papers met the stringent criteria and were subjected to detailed coding.

The review is structured around four research questions (RQs). RQ1 investigates the roles and contexts of XR. The majority of studies (27) use XR for “evaluation & testing” and 24 for “simulation,” while only a handful address training, prototyping, or robot‑robot collaboration. Application domains are split between “cross‑domain” fundamental HRI investigations (12 papers) and applied settings such as education (7) and healthcare (7). VR dominates the technology landscape (20 papers), followed by MR (9) and AR (5). Hardware is heterogeneous: HTC Vive/Pro (7), Microsoft HoloLens 2 (8), Meta Quest (4), but 12 papers do not specify the headset at all. Unity is the most common development platform (19), with occasional use of ROS, custom engines, or Unreal. The virtual robot platforms are primarily Pepper (8) and NAO (6), with several studies employing fully virtual agents. A notable 23 of the 33 papers rely on Wizard‑of‑Oz control to generate speech or gestures, indicating that the robots are often passive visual stimuli rather than autonomous agents. Interaction types are limited: 15 studies report no participant‑robot interaction, 10 use voice commands, and only a few allow free‑form dialogue or object handling.

RQ2 examines data collection and analysis methods. Self‑report questionnaires (29) and observational coding (25) are the dominant data sources; interviews appear in only 10 papers, and physiological measurements (e.g., heart rate, skin conductance) are present in just 7. Despite the built‑in eye‑tracking, pupillometry, and cognitive load sensors of modern head‑mounted displays, real‑time biometric metrics are scarcely used. All studies report descriptive statistics; 27 employ inferential tests, 16 use comparative statistics, 10 report correlations, and 9 incorporate qualitative analysis. This reliance on subjective measures and manual coding limits the richness of the data that XR could otherwise provide.

RQ3 identifies current challenges and proposes a four‑phase roadmap. The authors highlight (1) hardware limitations such as latency and resolution that affect ecological validity, (2) small, homogeneous participant samples (predominantly Western, young, male university students), (3) short experimental cycles that hinder longitudinal insights, and (4) insufficient reporting of XR specifications, which hampers reproducibility. To address these issues, the roadmap calls for: (i) strengthening application contexts (e.g., education, healthcare, public spaces) to move beyond laboratory‑only scenarios; (ii) establishing robust, testable technological iterations with transparent reporting of hardware, software, and performance metrics; (iii) embedding diversity in research teams and participant pools to improve external validity; and (iv) developing reporting standards, such as a unified taxonomy for XR roles, hardware, and robot characteristics.

RQ4 focuses on the demographics of researchers and participants. The research teams are largely composed of computer‑science or engineering scholars, with a strong male bias (≈70 %). Institutional affiliations are concentrated in Germany, the United States, and other Western nations. Participant demographics mirror the research teams: most are aged 20‑30, enrolled in higher‑education programs, and skew male. This lack of diversity raises concerns about the generalizability of findings to broader populations.

Overall, the paper concludes that while XR offers unparalleled control, immersion, and the potential for ecologically valid social HRI experiments, its current use is constrained by methodological narrowness, under‑exploited sensor capabilities, and demographic homogeneity. By following the proposed roadmap—expanding real‑world application domains, rigorously documenting and iterating XR technology, prioritizing diverse samples, and adopting standardized reporting—XR can evolve from a laboratory prototype into a mature, reliable research instrument for studying social robots in realistic settings.