ROBSO-PM: Robots for Museum Inclusion

• ROBSO-PM Project is a research initiative that integrates social and telepresence robots to address accessibility and inclusion challenges in small museums through adaptive storytelling and cultural mediation.
• The project employs a multimodal approach, combining personalized narratives, multilingual content, and co-designed interaction workflows to bridge sensory, cognitive, and cultural gaps.
• It leverages diverse case studies in both urban and mountain museums to validate its framework, aiming for future transferability to under-resourced cultural institutions.

Searching arXiv for the cited project paper and closely related supporting papers. arXiv search query: (Balossino et al., 8 Aug 2025) ROBSO-PM Social and Telepresence Robots for Accessibility and Inclusion in Small Museums The ROBSO-PM Project is a research initiative concerned with accessibility and inclusion in small museums through the use of social robots and social telepresence robots. Its stated aim is to improve access for on-site visitors, including foreign and disabled visitors, and for remote participants with limited mobility who cannot physically visit museums. The project is centered on three case studies—the Museum of the Holy Shroud in Turin, the Museum of the Champlas du Col Carnival, and the Pragelato Museum of Alpine Peoples’ Costumes and Traditions—and frames robotics not only as a delivery mechanism for information but as a form of cultural mediation organized around storytelling, robot personality, empathy, personalization, and, in telepresence use, explicitly defined roles and autonomy between the robot and the remote participant (Balossino et al., 8 Aug 2025).

1. Project scope and rationale

ROBSO-PM addresses a class of museums that are often underexamined in museum robotics: small institutions, especially those located in low-density population areas. The project identifies accessibility barriers as perceptual, cultural, and cognitive. Perceptual barriers include the absence of services for sensory disabilities such as vision and hearing impairments; the proposed mitigations include multilingual spoken explanations, text alternatives, sign language, and multimedia content. Cultural barriers arise when foreign visitors or culturally diverse audiences encounter language and mediation difficulties; in this framework, the robot is positioned as a cultural mediator able to provide language support and context-rich storytelling. Cognitive barriers concern the limited availability of simplified explanatory content and adaptive guidance in small museums; the project therefore foregrounds personalization and empathy to support visitors with cognitive limitations through tailored, understandable narratives (Balossino et al., 8 Aug 2025).

A distinctive feature of the project is its territorial emphasis. Accessibility shortcomings are presented as especially evident outside urban centers, where mountain museums may face seasonal constraints and limited resources. Telepresence is therefore framed not only as a mobility accommodation but also as a strategic response to geographic and demographic marginality. The expected functions of telepresence include enabling remote visits for people in nursing homes or isolated communities, facilitating reconnection with places of origin, promoting lesser-known destinations, enabling year-round access, and countering overtourism while amplifying cultural participation where physical footfall is low (Balossino et al., 8 Aug 2025).

The project description is prospective rather than retrospective. It reports objectives, planned workflows, and anticipated impacts, not completed outcome studies. A common misunderstanding would be to read ROBSO-PM as a report of already validated gains in inclusion; the source instead presents a project architecture and research agenda whose empirical verification is still pending (Balossino et al., 8 Aug 2025).

2. Case-study museums and contextual constraints

The three case studies are heterogeneous in profile, which is central to the project’s design logic. The Museum of the Holy Shroud in Turin is described as a small but globally known institution. Its audience is therefore understood to include foreign visitors requiring multiple language options, along with visitors drawn by a high-profile artifact associated with religious, historical, and scientific narratives. This setting implies a need for robust multilingual capabilities, sensitive cultural mediation, and support for varied visitor profiles, including older visitors and faith-based tourism segments. Although urban, the museum remains small, so space constraints and safety in crowded or narrow areas remain relevant (Balossino et al., 8 Aug 2025).

The two mountain museums introduce a different operational regime. For the Museum of the Champlas du Col Carnival, the available description supports only a partial characterization: a plausible implication is that its collections concern local carnival traditions and community memory. The project summary associates this setting with seasonal flows, limited staff, and visitors interested in local heritage. In this context, robots are expected to provide consistent interpretive services even during off-peak periods, while telepresence may extend access to former residents, schools, or cultural groups. The intended content design emphasizes ethnographic storytelling, rituals, costumes, and visual supports adapted for cognitive accessibility (Balossino et al., 8 Aug 2025).

The Pragelato Museum of Alpine Peoples’ Costumes and Traditions is similarly framed as an ethnographic and material-culture environment. A plausible implication is that its visitors may include families, students, and local communities, and that empathy and personalization are important for linking everyday alpine traditions to contemporary relevance. The project summary further highlights typical small-museum constraints—small spaces, potential connectivity limitations, and minimal staffing—which favor solutions that are easy to operate and maintain and that can function autonomously while still fitting staff workflows (Balossino et al., 8 Aug 2025).

Taken together, the three museums function less as interchangeable deployment sites than as differentiated testbeds for the same inclusive-design agenda. This suggests that ROBSO-PM is oriented toward transferability through contextual adaptation rather than through a single uniform interaction template.

3. Accessibility model and inclusive museum mediation

The project’s accessibility model is explicitly multimodal and audience-adaptive. For perceptual accessibility, the proposed modalities include multilingual spoken explanations, text alternatives such as captions or on-screen text, sign language as intended support, and multimedia content. For cultural accessibility, the robot is conceived as an intermediary capable of bridging language and interpretive gaps through narrative mediation. For cognitive accessibility, the emphasis falls on simplified explanatory content, adaptive guidance, and tailored narratives that remain understandable without flattening the museum’s interpretive richness (Balossino et al., 8 Aug 2025).

This combination is significant because ROBSO-PM does not treat accessibility as a single impairment-centered engineering problem. Instead, it joins sensory accommodation, linguistic mediation, and cognitive adaptation within one interactional framework. The source repeatedly links these functions to human-centered design and co-design, indicating that accessibility is to be specified with stakeholders rather than imposed as a fixed technical package. The same logic applies to remote access: telepresence is not presented merely as a substitute for presence, but as an accessibility modality that expands participation for people who are immobile, geographically distant, or excluded by the seasonal and infrastructural limits of small museums (Balossino et al., 8 Aug 2025).

The project also associates accessibility with inclusion in a broader cultural-policy sense. By extending access to under-served publics, enabling year-round participation, and supporting smaller institutions in low-density territories, the intervention is described as contributing to economic, cultural, and social development. The available account does not quantify these effects empirically, but it does define them as intended outcomes of the framework (Balossino et al., 8 Aug 2025).

4. Robotic applications, tasks, and operational workflows

ROBSO-PM organizes its robotic intervention around two principal applications. The first is the social robot guide for inclusive on-site visits. In this role, the robot is expected to guide visitors through exhibits, provide information and orientation, deliver multilingual spoken content, offer text alternatives, potentially support sign language and multimedia, and adapt content to visitor needs, including foreign language requirements and sensory or cognitive access needs. The scenarios described include robot-led tours for small groups or individuals, ad hoc assistance for foreign visitors, and accessibility-oriented tours using simplified language and rich audio descriptions for visually impaired users (Balossino et al., 8 Aug 2025).

The planned workflow for this guide role is structured but not algorithmically specified. It includes co-designed scripts and routes, autonomous navigation or movement with planned stops, dialogue initiation, triggering of multimedia content, interaction monitoring, and fault detection with requests for assistance when needed. The corresponding success criteria are also defined prospectively: increases in inclusive participation, visitor engagement metrics, interaction logs showing sustained attention, perceived empathy scores measured with the Perceived Empathy of Technology Scale, positive satisfaction surveys, and higher visit frequency among target groups such as foreign visitors, disabled visitors, and students (Balossino et al., 8 Aug 2025).

The second application is the social telepresence robot for remote access. Here the robot is intended to enable remote participants—for example, people in nursing homes or isolated areas—to experience the museum, interact with on-site visitors and staff, move through the museum safely, present content, and maintain its own identity and social guidance role. The project explicitly distinguishes this model from conventional telepresence in which the robot is only an embodied extension of a remote operator. Instead, ROBSO-PM aims to preserve the robot as an active cultural mediator with defined autonomy and a recognizable social role (Balossino et al., 8 Aug 2025).

The telepresence workflow includes booking and scheduling through an application ecosystem, on-site preparation by museum staff, hybrid control and shared autonomy, content delivery and social interaction during the visit, and post-visit data collection through logs and feedback. Planned success criteria include increased telepresence usage, high perceived empathy and usability, evidence of social presence and engagement for remote visitors, positive feedback from staff and local participants, and safe navigation with minimal incidents. The project summary therefore treats telepresence not as a transport layer but as a collaborative sociotechnical arrangement among robot, remote participant, and museum personnel (Balossino et al., 8 Aug 2025).

5. Storytelling, robot personality, empathy, and personalization

The project identifies storytelling, robot personality, empathy, and personalization as central research topics rather than as peripheral interface refinements. In small museums, narrative is presented as especially important for connecting local heritage to visitors’ experiences. Accordingly, the robot is expected to “offer its own perspective, physical presence, and kinetic expressiveness,” which implies an embodied and performative narrative style rather than purely informational output (Balossino et al., 8 Aug 2025).

Although the narrative structure is not formally specified, the intended format consists of guided sequences organized as routes and stops with contextualized interpretations tailored to audience needs. This can include multilingual narration for foreign visitors and simplified explanations for cognitive accessibility. The robot personality is defined as a distinctive social identity of Sanbot that persists across both on-site and telepresence modes. Its operationalization is to be co-designed with museum stakeholders and visitors. The available summary suggests that personality may involve tone, pacing, gesture repertoire, and empathic behavior, but these details remain to be determined through later project phases (Balossino et al., 8 Aug 2025).

Personalization is described as a central research challenge, yet the project does not specify algorithms, heuristics, formal models, or equations for visitor adaptation. The likely inputs named in the project phases and evaluation plan are questionnaires, interviews, focus groups, interaction logs, and satisfaction surveys. Group modeling is referenced in order to support multi-user adaptation, and future work names stereotypical user and group modeling and empathic interaction as directions for development. This suggests a progression from qualitative requirement elicitation toward increasingly structured adaptation logic, though the project paper does not yet define that logic technically (Balossino et al., 8 Aug 2025).

A related misconception concerns empathy. In ROBSO-PM, empathy is not treated as a vague anthropomorphic claim; it is an evaluable design dimension to be assessed through validated instruments such as the Perceived Empathy of Technology Scale. At the same time, the absence of implementation detail means that the exact behaviors through which empathy will be instantiated are still open research questions (Balossino et al., 8 Aug 2025).

6. Platform choice, evaluation framework, transferability, and open challenges

Sanbot is the chosen robot platform for all three museums. The project paper does not specify the Sanbot model, hardware configuration, sensors, middleware, perception stack, or speech and NLP pipeline. It also does not provide detailed locomotion or software-stack information. Instead, it refers to general museum-robot requirements drawn from prior systems, including collision avoidance, emergency stop buttons, autonomous movement planning, fault detection, and cognitive and interactive capabilities. The project further emphasizes multimodal communication and adaptability, taking inspiration from recent advances such as IIT’s R1 tested at the GAM (Balossino et al., 8 Aug 2025).

The architecture described in the source is therefore conceptual rather than implementation-complete. Its implied components include movement and navigation with autonomous planning and adaptation in complex environments; dialogue and storytelling modules supporting empathy, personalization, and multimodal outputs; a telepresence application ecosystem for scheduling, session management, and streaming; monitoring and logging for interaction analysis; and co-design interfaces for iterative development and testing. Privacy safeguards, communication protocols, codecs, remote-user interface details, and specific autonomy levels are not specified. Because telepresence, personalization, and logging are all present, privacy, consent, and data protection emerge as necessary design requirements, but the paper does not yet detail them (Balossino et al., 8 Aug 2025).

The evaluation methodology combines formative and summative assessments with mixed methods. Quantitative measures include interaction logs, visit frequency, telepresence usage, and social media analytics, while qualitative measures include satisfaction surveys and interviews with museum managers. The Perceived Empathy of Technology Scale is the one validated instrument explicitly named. The paper estimates beneficiaries as 17,000 general visitors, 2,400 people with disabilities, 6,800 foreign visitors, and 2,000 students, but it does not provide a concrete study sample size, sampling strategy, statistical analysis plan, effect sizes, confidence intervals, or significance tests. No comparative results across the three museums are yet available, because the project description reports expected outcomes rather than completed measurements (Balossino et al., 8 Aug 2025).

The anticipated impacts include enhanced accessibility across mobility, sensory, and cognitive dimensions; increased visitor engagement and inclusive participation; attraction of new visitors, including foreign visitors and younger audiences; and, for mountain museums, year-round visits and broader reach through telepresence. The project also explicitly aims to produce a transferable framework for other small museums based on inclusive design principles and human-centered co-design. Best practices implied by the design include early requirement analysis with stakeholders through questionnaires, interviews, and focus groups; co-design and iterative testing before and after release; integration of storytelling and robot personality to sustain engagement; careful handling of safety and navigation in constrained spaces; and ongoing monitoring through quantitative logs, qualitative feedback, and validated empathy and usability scales (Balossino et al., 8 Aug 2025).

Open challenges remain substantial. These include operationalizing robot personality and empathy in ways that demonstrably improve accessibility and inclusion; designing a telepresence collaboration model that balances the robot’s identity and guide role with meaningful remote-user agency; handling ambient noise, short interaction times, uncooperative users, spatial constraints, staffing limits, and connectivity limitations typical of small museums and mountain locations; and measuring inclusion outcomes rigorously across diverse user groups while meeting privacy and ethical standards. Future directions explicitly named for the project are IoT integration, stereotypical user and group modeling, empathic interaction, and multimodal design strategies (Balossino et al., 8 Aug 2025).