Introduction

The rapid advancement of digital technologies has revolutionized how cultural heritage is preserved, accessed, and experienced. Cultural heritage, which includes both tangible elements such as monuments, artifacts, and textiles, as well as intangible aspects like oral traditions, rituals, and craftsmanship, forms the backbone of a community’s identity and legacy. However, these heritage assets face increasing risks from physical degradation, loss of traditional knowledge, and limited accessibility, especially for younger generations (Tešin et al., 2021). Innovative digital solutions are therefore required to safeguard and promote heritage in ways that transcend geographical and generational boundaries. Among emerging technologies, the metaverse and generative artificial intelligence (AI) have shown transformative potential for reimagining how people experience, learn about, and emotionally connect with heritage.

The metaverse is a persistent, shared, and immersive virtual environment that merges physical and digital realities through technologies such as virtual reality (VR) and augmented reality (AR). Within this environment, cultural heritage, including tangible artifacts and intangible traditions—can be digitally reconstructed, preserved, and transmitted in innovative ways (Buragohain et al., 2024). This approach aligns with international frameworks such as UNESCO’s Charter on the Preservation of Digital Heritage (UNESCO, 2009) and the PERSIST Project (UNESCO, 2015), which emphasize the importance of inclusive access, long-term preservation, and collaborative stewardship of digital cultural assets. The motivation for adopting AI-integrated metaverse systems in museums arises from their potential to revitalize public engagement, transform static exhibitions into interactive narratives, and bridge traditional knowledge with digital learning environments. By allowing visitors to interact with digitized artifacts, explore 3D reconstructions, and converse with intelligent virtual assistants, museums can create immersive experiences that foster empathy, curiosity, and a deeper understanding of cultural identity. For example, users can explore 3D models of historical sites, engage with traditional practices, or interact with digitized artifacts in real time (Kouroupi and Metaxas, 2023). The integration of generative AI a form of AI capable of autonomously creating novel content such as narratives, visuals, and dialog further enhances this potential. Generative AI allows for context-aware storytelling and adaptive personalization, making cultural heritage experiences more dynamic and engaging. This is especially valuable for preserving intangible cultural heritage, such as rituals and craftsmanship, through embodied interaction and immersive digital storytelling (Chamola et al., 2024; Hutson, 2024a).

Despite advancements in digital technologies, research on the application of generative AI and the metaverse in cultural heritage remains limited. Many virtual heritage projects rely on static representations or pre-programmed interactions, which limit user engagement and fail to leverage the full potential of emerging technologies. For instance, Huggett (2020) notes that virtual heritage reconstructions often emphasize visual accuracy over interactivity, resulting in passive experiences. Similarly, Ribeiro et al. (2024) and Spennemann (2023) highlight that while VR and AR are increasingly used in museum contexts, their integration with adaptive AI features remains underdeveloped, with most systems offering scripted or limited-choice interactions. These gaps highlight the need for a stronger theoretical and practical rationale for employing AI in heritage environments, particularly in contexts like Thailand, where intangible craftsmanship and local identity form the essence of cultural sustainability. This study, therefore, focuses on integrating generative AI as a virtual assistant within a metaverse framework to enhance contextual guidance, ethical storytelling, and personalized learning. By providing context-aware guidance, immersive storytelling, and personalized learning experiences, generative AI can preserve the authenticity of cultural assets while engaging diverse audiences. The case study of Lamphun Brocade Fabric a traditional textile from Northern Thailand illustrates these challenges and opportunities, including declining weaving knowledge and limited global awareness. Data collected from the Wieng Yong House Museum, which holds fabrics over 100 years old, serves as a foundation for creating a generative AI-integrated metaverse that promotes both heritage preservation and community empowerment.

Related work on the metaverse of cultural heritage

The term metaverse refers to a shared, virtual, and persistent digital space that integrates physical reality with virtual and augmented environments (Ng, 2022; Zhou et al., 2024), often leveraging advanced technologies such as VR, AR, and AI (Ritterbusch & Teichmann, 2023). It provides users with immersive, interactive experiences that enable real-time communication, exploration, and engagement. Within the context of cultural heritage, the metaverse serves as a digital platform to preserve, promote, and reimagine both tangible and intangible cultural assets. By creating immersive environments, the metaverse offers novel ways for users to explore historical narratives, interact with cultural artifacts, and gain deeper insights into diverse traditions and practices, transcending geographical and temporal boundaries. This transformative approach has been demonstrated in various applications, including the digital preservation of historical figures (Buragohain et al., 2024), the promotion of intangible cultural traditions (Huggett, 2020), and the use of immersive environments for interactive learning and tourism (Baker et al., 2023).

The metaverse has been explored extensively for its potential in preserving and promoting cultural heritage through a variety of innovative projects. In Amiais, Portugal, a metaverse platform was co-designed to promote cultural heritage and tourism by integrating storytelling and virtual exploration. Similarly, in China, the metaverse has been employed to preserve the narratives of historical figures such as Zhu Xi from the Song Dynasty, creating immersive experiences that document and showcase these stories for a global audience (Fan et al., 2022). In Ivrea, Italy, a VR platform was developed to commemorate the intangible cultural heritage of the Italian Resistance, emphasizing local traditions and history through engaging, interactive environments (Innocente et al., 2024). Furthermore, in Singapore, a metaverse focused on the Eight Immortals sculptures at Haw Par Villa showcased the ability to digitally promote Taoist heritage (Zhang et al., 2022). Other notable projects include the use of the ZEPETO platform for youth education on the history and architecture of Pingyao Ancient City in China, demonstrating improved engagement and learning outcomes (Gao and Lee, 2024), and a metaverse initiative for Babylon, Iraq, that explores virtual accessibility and immersive preservation for cultural tourism (Wu et al., 2022). These projects underscore the metaverse’s versatility in safeguarding both tangible and intangible cultural heritage while fostering education, tourism, and cultural appreciation. The overview of related work on the metaverse of culture heritage is shown in Table 1.

Table 1 The overview of related work on the metaverse of cultural heritage.
Full size table

Intangible cultural heritage in Thailand

Thailand possesses a rich and diverse intangible cultural heritage that reflects community identity, local wisdom, and artistic craftsmanship. Intangible heritage includes oral traditions, performing arts, rituals, social practices, and traditional craftsmanship. In Northern Thailand, particularly in the Lamphun and Chiang Mai provinces, hand-woven textiles such as Lamphun Brocade Fabric (locally known as pha yok Lamphun) represent both artistic mastery and communal identity (Chai-Arayalert et al., 2021). These textiles embody Buddhist symbolism, cosmological motifs, and social meanings that are passed down through generations of female artisans. However, modernization, industrial production, and generational gaps have contributed to the decline of traditional weaving knowledge (Sombat & Mahavarakorn, 2022). Existing preservation efforts mainly focus on documentation or museum displays rather than interactive or AI-enhanced approaches, leaving significant opportunities for applying immersive technologies such as the metaverse and generative AI to revitalize community-based craftsmanship and sustain local heritage economies.

Artificial intelligence in cultural heritage

AI has become an essential tool in the preservation, promotion, and reimagination of cultural heritage. AI technologies enable the digitization and safeguarding of both tangible assets, such as artifacts and monuments, and intangible heritage, including oral histories, traditions, and performances (Harisanty et al., 2024). Generative AI, in particular, has revolutionized the field by creating digital replicas of cultural artifacts, reconstructing lost or damaged pieces, and simulating historical designs. For instance, generative AI tools have been successfully applied to co-create cultural artifacts and enhance storytelling, fostering emotional connections and engagement with heritage (Fu et al., 2024). AI-powered models also process vast amounts of historical data to ensure that cultural knowledge is preserved, accessible, and disseminated globally (Ghaith and Hutson, 2024).

Beyond preservation, AI fosters deeper engagement through immersive and interactive experiences. AI-driven virtual assistants and chatbots enhance cultural metaverses by offering real-time, context-aware narratives tailored to individual users, transforming educational and exploratory experiences. Generative AI takes this further by dynamically creating content, such as personalized stories and simulations that adapt to user interactions. For example, generative storytelling paired with AI-visual illustrations has been employed to promote cultural tourism, creating engaging experiences that combine education and entertainment (Ferracani et al., 2024). Additionally, AI has enabled the development of interactive challenges, such as quizzes and games, to foster participation and learning in museums and cultural heritage sites (Ribeiro et al., 2024). Despite its transformative potential, AI in cultural heritage presents challenges, including ethical concerns surrounding data privacy (Spennemann, 2023), cultural sensitivity, and the mitigation of bias in AI models (Foka and Griffin, 2024). These developments align with recent theoretical perspectives in human–computer interaction and digital heritage preservation, which emphasize immersive interaction, participatory design, and socio-technical systems frameworks as foundations for sustaining long-term user engagement and cultural learning (Lu et al., 2025; Ming et al., 2024).

However, there is still a lack of research on implementing generative AI as a virtual assistant in the metaverse for dynamic, immersive, and personalized engagement in cultural heritage preservation and education. Current virtual assistants and AI applications in the metaverse often rely on pre-programmed responses and are primarily focused on education in medical and engineering contexts (Chheang et al., 2023). By implementing generative AI as a virtual assistant in the metaverse, this research could provide a deeper understanding of how generative AI can be utilized specifically in the cultural heritage context. This advancement would demonstrate how AI-driven personalization and adaptability can transform cultural heritage engagement, education, and preservation for future generations. In particular, this study extends current metaverse research by situating generative AI within a culturally grounded case, the Lamphun Brocade Fabric, to explore how intelligent systems can support the preservation of intangible craftsmanship, ethical storytelling, and sustainable cultural revitalization in Thailand.

Project aim and objectives

The project was a collaboration with [University Name] and the Lamphun Weaving to promote Lamphun Brocade Fabric as a valuable cultural heritage asset. To achieve this goal, the project aims to leverage generative AI and metaverse technologies to preserve and promote the cultural heritage of Lamphun Brocade Fabric. By integrating cutting-edge digital tools, the project seeks to address the challenges of cultural preservation and the declining traditional knowledge associated with Lamphun weaving practices. Specifically, the project aims to:

  • Create a virtual metaverse environment for global users to explore and engage with the cultural heritage of Lamphun Brocade Fabric.

  • Digitally preserve tangible and intangible elements, including artifacts, weaving techniques, and cultural knowledge.

  • Support artisans through a digital marketplace to promote and sell brocade products globally, fostering sustainability.

  • Enhance education by providing tools to study the historical and cultural significance of Lamphun Brocade Fabric.

Achieving these objectives will bring significant benefits to the preservation and promotion of Lamphun Brocade Fabric. The virtual metaverse environment will connect global audiences with Lamphun’s rich cultural heritage, while digital preservation safeguards traditional knowledge for future generations. The digital marketplace will empower local artisans by providing access to global markets and fostering economic sustainability. Additionally, educational tools will enhance awareness and appreciation of Lamphun’s historical and artistic significance, ensuring the enduring legacy of this cultural heritage.

Theoretical framework

This study is grounded in four complementary theoretical and methodological foundations. From human–computer interaction, it adopts presence theory to explain how immersive environments elicit realistic responses and sustain user engagement (Slater et al., 2022). From museum learning, the research draws on the Contextual Model of Learning, which conceptualizes learning as the dynamic interplay of personal, sociocultural, and physical contexts, an approach that informs the project’s use of adaptive, visitor-centered narratives and participatory interaction (Falk & Dierking, 2005). Regarding digital-heritage methodology, the project aligns with international standards emphasizing transparency, authenticity, and methodological rigor in computer-based visualization. Finally, in addressing intangible cultural heritage (ICH), the framework follows the Convention for the Safeguarding of the Intangible Cultural Heritage, emphasizing transmission, community participation, and respect for cultural values. Collectively, these frameworks ensure that the generative-AI-integrated metaverse developed in this study is technologically innovative, pedagogically sound, ethically responsible, and culturally grounded.

Metaverse framework design

The metaverse framework for preserving and promoting Lamphun Brocade Fabric integrates advanced technological systems into a cohesive design aimed at cultural heritage preservation and engagement (Hutson, 2024b). The framework comprises three interrelated systems: the Heritage Base System, the Immersive Interaction System, and the Technology and Engagement System. Each component plays a critical role in offering a holistic approach to integrating generative AI and immersive virtual environments. The Heritage Base System collects cultural heritage data, encompassing both intangible and tangible elements, to ensure data integrity and knowledge preservation. The Immersive Interaction System enhances user engagement by facilitating realistic cultural experiences through 3D modeling and virtual reality environments. Meanwhile, the Technology and Engagement System incorporates AI and gamification strategies to captivate users and sustain interest. These systems are designed based on prior applications of Generative AI in the metaverse (Chamola et al., 2024; Lifelo et al., 2024), with the goal of promoting Lamphun Brocade Fabric textiles. An overview of the design is illustrated in Fig. 1.

Fig. 1
figure 1

The metaverse design for Lamphun Brocade Fabric.

Full size image

Heritage Base System

The Heritage Base System is designed as the core layer of the foundational structure of the metaverse, created to collect and manage digital data for the preservation, authenticity, and accessibility of cultural information related to Lamphun Brocade Fabric. This system is critical for maintaining the integrity of the digital ecosystem and serves as a robust repository for both tangible and intangible cultural heritage. The two key components of this system are the Metadata Integrity System and the Knowledge Cultural Repository.

Metadata Integrity System

The Metadata Integrity System is designed to collect a structured set of descriptive elements that define the content, context, and structure of an archive (Alma’aitah et al., 2020), ensuring the quality, reliability, and authenticity of the digital data stored within the metaverse. This system adheres to international standards and best practices to maintain consistency and trustworthiness in the documentation of cultural heritage. To standardize metadata, the VRA Core (Visual Resource Association) standard (Mixter, 2014) and Dublin Core (Giannoulakis et al., 2018), both widely used for cultural heritage applications, were selected. Furthermore, the system incorporates Data Exchange Protocols to enable the secure and efficient sharing of cultural data across platforms, including servers, websites, and virtual reality environments. JSON (Pezoa et al., 2016) was chosen as the protocol for data interchange due to its efficiency and compatibility. The cloud-based storage system collects and maintains the integrity of cultural data, further strengthening the credibility of the metaverse as a reliable tool for cultural preservation and engagement. An overview of the technology architecture for the heritage system is shown in Fig. 2.

Fig. 2
figure 2

An overview of the technology architecture for the heritage system is shown in Fig. 2.

Full size image

Knowledge Cultural Repository

The Knowledge Cultural Repository serves as a comprehensive archive that preserves and organizes the rich cultural heritage associated with Lamphun Brocade Fabric by collecting both tangible and intangible elements of the culture and visualizing them within the metaverse. Using the VRA Core and Dublin Core metadata standards, the repository encompasses Digital Cultural Archives, which store high-resolution images, detailed descriptions, and 3D models of artifacts, providing a digital representation of Lamphun’s weaving traditions for global exploration and learning. It focuses on preserving Tangible Cultural Knowledge, including traditional weaving tools, intricate patterns, and unique fabric designs, as well as Intangible Cultural Knowledge, such as oral traditions, historical narratives, and weaving techniques. An overview of the digitalization methods and the Knowledge Cultural Repository is shown in Fig. 3.

Fig. 3
figure 3

An overview of the digitalization methods and the knowledge cultural repository.

Full size image

Tangible and intangible knowledge

Cultural knowledge of Lamphun Brocade Fabric encompasses both tangible and intangible elements, each essential for preserving its rich heritage. Tangible Knowledge refers to physical aspects such as traditional weaving tools, intricate fabric patterns, antique weaving machines, and artifacts like furniture and pottery. These are meticulously documented using various digital methods (Storeide et al., 2023), including 3D modeling, high-resolution scanning, and photogrammetry techniques (Kasapakis et al., 2018). These digital archives provide detailed, interactive visualizations of physical objects in the metaverse, ensuring global accessibility. A summary of tangible objects is shown in Table 2. On the other hand, Intangible Knowledge comprises non-physical elements, including oral traditions, historical narratives, weaving techniques, and ritual practices, preserved through video recordings, audio documentation, and textual descriptions to ensure the continuity of practices and skills passed down through generations. Storytelling, interviews with artisans, and demonstrations of weaving techniques are digitized to create immersive experiences that capture the essence of the cultural heritage. By integrating these elements into structured, accessible digital formats, both tangible and intangible aspects of Lamphun’s cultural heritage are securely stored in the cloud-based Knowledge Cultural Repository. An example of tangible and intangible knowledge representations is shown in Figs. 4 and 5.

Table 2 Summary of digitized objects in tangible cultural knowledge.
Full size table
Fig. 4
figure 4

Intangible knowledge of weaving techniques by simulation.

Full size image
Fig. 5
figure 5

Example of ancient Lamphun Brocade Fabric in museum.

Full size image

Immersive Interaction System

The Immersive Interaction System is the core layer that provides an immersive experience for users in the metaverse. This system bridges the gap between digital cultural knowledge and user interaction, offering an interactive platform that enables users to explore, learn, and connect with the cultural heritage of Lamphun Brocade Fabric. The system consists of two primary modules: Immersive Virtual Representation and Interactive User Experience, each contributing to the depth and accessibility of the metaverse.

Immersion Virtual Representation

The Immersive Virtual Representation module focuses on creating authentic and visually rich environments that reflect the cultural heritage of Lamphun Brocade Fabric. This module utilizes advanced technologies, with Unity 2021.1a selected as the core graphic visualization software for the metaverse experience. We also created 3D reconstructions of the Wieng Yong House Museum, Wat Ton Kaew Temple, and Pagoda, which serve as landmark features of the metaverse, as shown in Fig. 6. Virtual environments are designed to replicate the settings and atmospheres of Lamphun’s cultural sites, with integrated lighting, sound effects, and spatial design to enhance sensory immersion. Additionally, dynamic animations bring digital representations to life by demonstrating weaving techniques, recreating historical events, and adding contextual depth to cultural artifacts.

Fig. 6
figure 6

General assistant (left) and focused tangible assistant (right) for AI interaction.

Full size image

Interactive User Experience

The Interactive User Experience module provides an engaging environment where users can actively interact with the metaverse, transforming it into a dynamic space for learning and exploration. This module incorporates a user-centered interface design (Reunanen et al., 2015) that prioritizes accessibility and ease of navigation, enabling users of all technical backgrounds to interact effortlessly with the content. Teleportation-based locomotion (Puritat et al., 2022) and navigation controls are being considered to provide realistic movement through virtual environments, replicating the experience of visiting real-world cultural heritage sites. To enhance interaction with AI-powered NPCs, the user interface includes two types of Assist Modes: General Assistant, which offers broad guidance and allows users to ask the AI questions via a voice microphone at any time, and Focused Tangible Assistant, which provides detailed interaction with specific cultural artifacts by requiring users to click on the artifact to access information. An example of these Assist Modes is shown in Fig. 6.

Technology and Engagement System

The Technology and Engagement System integrates advanced technologies and gamification strategies as the highest layer of the metaverse system to enhance user experience and engagement. This system consists of two main components: the Generative AI Integration and the Gamification Module. The Generative AI Integration module enables AI-driven interactions, including intelligent NPCs and storytelling modules, which assist users and answer questions based on their preferences. The Gamification Module incorporates interactive game elements, mini-games, and achievement/reward mechanisms to transform cultural learning into an entertaining and rewarding experience.

Generative AI Integration

The Generative AI Integration module leverages advanced technologies to deliver a seamless and immersive user experience within the metaverse, particularly through virtual assistants in virtual reality environments (Chheang et al., 2023; Krauss et al., 2024). The system architecture incorporates ChatGPT-4o for natural language processing, enabling dynamic and context-aware interactions, while OpenAI’s Speech-to-Text API processes voice inputs, and Whisper provides text-to-speech conversion for natural and fluent communication. Unity3D powers 3D visualization, creating lifelike virtual environments, and a MySQL database stores exhibit metadata to ensure real-time access to cultural information. To ensure cultural specificity, the generative model (ChatGPT-4o) was customized using locally sourced datasets that included expert interviews, historical documentation, museum exhibit metadata, and traditional narratives related to Lamphun Brocade Fabric. To avoid misinformation and biased data from the AI, we defined clear boundaries for AI responses based on the validated dataset and evaluated the outputs in consultation with local stakeholders. This approach allowed the AI assistant to provide contextually accurate, culturally grounded responses tailored to the heritage content. Together, these technologies facilitate real-time, interactive responses as users explore the metaverse, fostering a deeper connection to the cultural heritage of Lamphun Brocade Fabric. The design of the Generative AI Integration system is illustrated in Fig. 7.

Fig. 7
figure 7

Overview of the generative AI integration system design.

Full size image

Gamification Module

The Gamification Module provides the benefits of positive psychology to the metaverse by incorporating engagement, flow experiences, alternate reality experiences, and hedonic experiences (Cha et al., 2024; Thomas et al., 2023). The module is designed to include gamification features such as game elements and mini-games that challenge users to replicate traditional weaving patterns, solve cultural puzzles, or complete quests related to Lamphun Brocade Fabric. An achievement system is integrated, enabling users to earn points, badges, and rewards for completing activities and milestones. Additionally, interactive game elements, including leaderboards and mini-games, foster a sense of competition and community among users while assessing their knowledge and encouraging participation. By blending education with entertainment, the Gamification Module transforms learning about Lamphun Brocade Fabric into an enjoyable and immersive experience. The gamification design applied to the system is illustrated in Fig. 8.

Fig. 8
figure 8

Mini-games quiz (left) and achievement system (right) in the gamification module.

Full size image

System evaluation

Regarding our generative AI-integrated metaverse for preserving and promoting the cultural heritage of Lamphun Brocade Fabric, we successfully developed an immersive metaverse system that incorporates advanced technologies, including generative AI and a gamification module. The generative AI component focuses on personalizing the user experience by providing assistance during visits, while the gamification module is designed to enhance user engagement and enjoyment. To evaluate the system, we employed a descriptive research method using a cross-sectional survey design, collecting post-interaction data from participants to assess usability, engagement, and overall user satisfaction.

Participants

To evaluate the metaverse, we invited student volunteers via social media who consented to provide feedback on our system. A comprehensive study was conducted with 85 participants, consisting of 62 bachelor’s degree students, 15 master’s degree students, and 8 doctoral students from [University Name]. The volunteers were recruited from a variety of disciplines, including computer science, information studies, tourism, history, and engineering, to provide diverse perspectives on the system’s usability and engagement features. To ensure the well-being and comfort of the participants, only individuals with prior experience using virtual reality technology were invited. This criterion was implemented to minimize the risk of motion sickness or other health issues that can occur among users who are not accustomed to VR environments.

Survey questionnaire

To comprehensively evaluate the usability and user experience of the generative AI-integrated metaverse, a combination of three survey instruments was utilized: the System Usability Scale (SUS) (Brooke, 1996), the User Engagement Scale (UES) (O’Brien et al., 2018), and an open-ended questionnaire. The SUS, a standardized tool for assessing virtual reality systems (Huang and Lee, 2019), measured overall usability and user satisfaction through a 10-item survey rated on a 5-point Likert scale. The results provided valuable insights into both the learnability and usability aspects of the system. The UES assessed user engagement, focusing on key dimensions such as focused attention, usability, esthetic appeal, and perceived value, also employing a 5-point Likert scale for responses. Additionally, the open-ended questionnaire gathered qualitative feedback, enabling participants to share detailed reflections on their experiences, highlight the system’s strengths, and suggest areas for improvement. The questionnaire included two questions: (1) “How was your experience using the metaverse system?” and (2) “What were the strengths and weaknesses of the system?” These items encouraged participants to describe both their overall impressions and specific advantages or limitations of the system in their own words.

Equipment and procedure

The experiment was conducted at the Humanity Lab, [University Name], over a period of two weeks. The setup included two Meta Quest 3 virtual reality headsets and two high-performance PCs, each equipped with an Intel Core i7 processor, NVIDIA GeForce RTX 3080 Ti graphics card, and 32 GB of RAM. Participants were briefed on the study’s objectives, signed online consent forms, and were not provided with tutorials on using the virtual reality equipment or navigating the metaverse system. Each participant spent between 10 and 30 min interacting with the system, exploring its features, completing gamified tasks, and engaging with digital representations of cultural artifacts. During the experiment, two research assistants were present. One was responsible for managing the equipment, while the other guided participants and provided instructions for completing the questionnaires. After the session, participants completed a post-survey consisting of the System Usability Scale, User Engagement Scale, and an open-ended questionnaire, accessed through their own mobile devices using Google Forms. The collected data was analyzed to evaluate the system’s usability, engagement, and overall effectiveness. Feedback obtained from participants was subsequently used to refine the system’s AI response timing, improve the naturalness of the voice interface, and enhance the overall interaction design.

Data analysis

To analyze the data, both quantitative and qualitative methods were employed. Responses from the SUS were scored following the standard procedure by Brooke (1996), with odd-numbered items adjusted by subtracting 1 and even-numbered items scored by subtracting the response from 5; the total score was then multiplied by 2.5 to produce a usability score ranging from 0 to 100. Descriptive statistics, including means and standard deviations, were calculated to evaluate overall usability. For the UES, participant responses were grouped into four dimensions: Focused Attention, Perceived Usability, Esthetic Appeal, and Reward, and analyzed using mean scores and standard deviations to assess engagement levels. The results were also visualized using stacked bar charts. Qualitative data from the open-ended questionnaire were analyzed using thematic analysis. Participant responses were manually reviewed, coded, and categorized into recurring themes. Frequencies and percentages were used to quantify the prevalence of each theme, allowing us to contextualize quantitative findings with detailed user insights.

Result

Result of System Usability Scale

To evaluate whether the metaverse achieved the expected usability and engagement, quantitative data from the System Usability Scale and User Engagement Scale were analyzed. The System Usability Scale score was calculated using responses to a 10-item questionnaire rated on a five-point Likert scale, with odd-numbered items scored by subtracting one from the response value and even-numbered items scored by subtracting the response value from five (Brooke, 1996), as shown in Table 3. The adjusted scores were summed and multiplied by 2.5 to produce a total score ranging from 0 to 100. SUS scores are categorized into five levels: F (0–59), D (60–69), C (70–79), B (80–89), and A (90–100) (Bangor et al., 2009; Huang and Lee, 2019), reflecting system usability. The metaverse achieved an average SUS score of 75.2, indicating a good level of usability and exceeding the industry software benchmark of 68. The overall SUS score is shown in Fig. 9.

Table 3 Result of System Usability Scale.
Full size table
Fig. 9
figure 9

Overview of system usability scale.

Full size image

Result of User Engagement Scale

The results of the User Engagement Scale, shown in Table 4 and Fig. 10, represent the key aspects of user experience within the metaverse. The highest-rated dimension, Reward, with a mean score of 4.14, reflects that participants felt a strong sense of accomplishment and satisfaction, driven by the well-designed gamification module. Similarly, Esthetic Appeal, scoring 4.01, emphasizes the positive reception of the metaverse’s visual design and interface, which contributed to an engaging and immersive experience. Perceived Usability, with a mean score of 3.97, further demonstrates that users found the platform efficient and easy to navigate, meeting the goal of accessibility and user-friendliness.

Table 4 Result of UES dimension.
Full size table
Fig. 10
figure 10

Result of stacked bar chart for system usability scale.

Full size image

However, the lower score for Focused Attention, with a mean of 3.45, suggests that participants experienced challenges in maintaining sustained concentration and immersion. A significant contributing factor may have been the instability of the AI system. Issues such as slow response times likely disrupted the flow of interaction, causing frustration and breaking user immersion (Cao et al., 2024; Pang et al., 2024). Additionally, inconsistencies in the tone of voice or communication style of the AI could have made interactions feel less natural or engaging (Huo et al., 2024), reducing trust and user satisfaction. Furthermore, technical inconsistencies or delayed responses from the AI assistant might have interrupted tasks (Wang et al., 2023), forcing users to shift focus or wait, which negatively impacted the overall experience. Insights from the open-ended responses could provide guidance on addressing these challenges and improving the system in future iterations.

Results of open-ended questionnaire

Regarding the results of the open-ended questionnaire, out of 68 participants, 51 (75%) provided responses to the open-ended questionnaire, while 17 (25%) did not. Thematic analysis of the qualitative responses revealed four major themes: (1) immersive environment and visual appeal, (2) system usability and interaction design, (3) gamification and learning engagement, and (4) areas for improvement and user concerns. Regarding the first theme, 40 participants (78.4%) praised the realistic visual environment, describing it as immersive, esthetically pleasing, and effective in conveying the cultural significance of Lamphun Brocade Fabric. For the second theme, 37 participants (72.5%) reported that the AI interface was easy to use, and 34 participants (66.7%) appreciated the simplicity and responsiveness of the locomotion controls. However, 7 participants (13.7%) commented that the museum environment felt intimidating, while 5 participants (9.8%) suggested improvements in object interactivity to enhance engagement within the metaverse space.

The third theme highlighted the system’s gamified learning experience. A majority of 42 participants (82.4%) responded positively to the mini-games and reward mechanisms, stating that they increased engagement and motivation to explore cultural content. Additionally, 9 participants (17.6%) suggested expanding the variety of gamified tasks to further enhance learning outcomes. The fourth theme reflected areas for improvement, particularly with the AI assistant. While 35 participants (68.6%) appreciated the AI’s personalized and context-aware support, 14 participants (27.5%) expressed frustration with delayed responses, robotic voice tone, and inconsistent communication. A total of 20 participants (39.2%) recommended enabling customization of the AI’s tone and personality to increase user comfort and relatability. Finally, 6 participants (11.8%), primarily female, raised hygiene concerns related to the shared VR equipment, suggesting that clear sanitation protocols be implemented to promote comfort and inclusivity in public or educational settings.

Discussion

User evaluation results and design insights

The results of the System Usability Scale, User Engagement Scale, and open-ended questionnaire experiment revealed valuable insights into the generative AI-integrated metaverse. The metaverse achieved an average SUS score of 75.2, indicating a good level of usability and aligning with findings from related research on cultural heritage software using virtual reality (Hulusic et al., 2023; Othman et al., 2021; Zilles et al., 2020). The User Engagement Scale emphasized the immersive experience provided by the metaverse, with Perceived Usability, Esthetic Appeal, and Reward receiving positive average scores of 3.97, 4.01, and 4.14, respectively, highlighting the effectiveness of the metaverse’s visual design and interface. These findings are attributed to the high quality of digitalization methods and realistic graphics, which significantly enhanced the overall user experience. However, Focused Attention scored the lowest, with a mean of 3.45, as participants reported difficulties in maintaining sustained concentration and immersion. Qualitative feedback identified issues with the AI system, including slow response times, an unnatural tone of voice, and occasional technical delays or inaccuracies. These issues disrupted the flow of interaction, caused frustration, and negatively affected user immersion. Additionally, hygiene concerns related to shared VR equipment were raised, emphasizing the importance of maintaining cleanliness to ensure a comfortable and inclusive user experience.

Based on our knowledge and experience in the development of the metaverse, several improvements can be implemented to address the identified problems and enhance the generative AI-integrated metaverse. First, AI customization should be prioritized to provide personalized interactions tailored to user profiles and preferences, creating a more engaging and relevant experience (Hatami et al., 2024). With the future trend of AI capacities expanding exponentially (Graylin and Rosenberg, 2024), improvements in processing speed are expected to eliminate lag times and delays, ensuring a seamless and responsive system. Additionally, based on our experience with frequent user interactions, common questions such as “What is the most interesting thing in this metaverse?” or “What is the most fun thing in this metaverse?” are often asked. Enabling the AI to proactively suggest answers to such frequently asked questions at the beginning of a user’s visit can significantly enhance the experience (Dwivedi et al., 2022). Lastly, issues related to the tone of voice, which are often dependent on API limitations, should be addressed by providing users with the ability to customize the AI’s tone and voice to match their preferences, further improving satisfaction and immersion. These enhancements would benefit researchers and museums aiming to create AI-powered NPCs for cultural heritage metaverses, collectively leading to a more intuitive, engaging, and user-centric metaverse experience.

Practical design considerations and AI ethics

In the broader context of AI and metaverse applications in cultural heritage, this project draws valuable lessons from prior initiatives while offering practical pathways for future implementation. Projects such as the Pingyao Ancient City and the Eight Immortals sculptures in Singapore have demonstrated how gamified storytelling and immersive environments can effectively engage younger audiences (Gao and Lee, 2024; Zhang et al., 2022). Similarly, the Zhu Xi metaverse highlights the potential of AI-driven narrative agents in delivering personalized, context-aware cultural education (Fan et al., 2022). These examples emphasize the importance of adaptive content, emotional engagement, and user-centered design. In response, the Lamphun project applies these lessons through practical features such as dynamic NPC tour guides, customizable AI voice and behavior, and modular gamification strategies tailored to specific cultural artifacts. Moreover, the use of standardized metadata frameworks and cloud-based repositories ensures long-term preservation and scalability. As a result, this project not only advances the digital preservation of Lamphun Brocade Fabric but also offers a transferable design model that can inform future practices among researchers, curators, and museums seeking to develop more inclusive, interactive, and sustainable AI-integrated metaverse experiences.

However, the integration of generative AI to support user interaction may lead to the generation of misinformation related to cultural knowledge, particularly when users pose questions beyond the scope of the training dataset. This raises serious ethical concerns, as without proper safeguards, such technologies risk misrepresenting cultural narratives, marginalizing community voices, or introducing algorithmic bias that undermines authenticity and trust (Fu et al., 2025; Spennemann, 2024; Tiribelli et al., 2024). These concerns are especially critical when AI-generated outputs are used to represent intangible heritage or historically underrepresented communities. Aligned with the principles outlined in the UNESCO Recommendation on the Ethics of Artificial Intelligence (UNESCO, 2023), which emphasizes inclusiveness, cultural diversity, and respect for data sovereignty, this project adopted proactive measures to ensure ethical implementation. Development was informed through ongoing consultation with local stakeholders, including artisans, curators, and cultural experts, to ensure that AI-generated content remained contextually accurate and respectful. Locally curated datasets were used to fine-tune the AI model, rather than relying on generalized or globally biased data, and clear boundaries were established for the AI system, including the ability to respond with “no information available” when users asked questions beyond its culturally validated dataset. These safeguards represent a practical implementation strategy that can guide other researchers, curators, and cultural institutions in adopting AI technologies in ways that amplify local voices and preserve cultural integrity.

Conclusion, limitations, and future work

This study aimed to design and develop a generative AI-integrated metaverse to foster visitors’ interest and curiosity in exploring the cultural heritage of Lamphun Brocade Fabric. The metaverse environment incorporates digitized representations of the fabric, interactive behaviors, and gamified features to enhance the learning experience and promote cultural knowledge. Following the system’s development, university students participated in evaluations through structured questionnaires. The evaluations using the System Usability Scale, User Engagement Scale, and open-ended questionnaires demonstrated that the metaverse effectively supports participants in terms of usability, immersion, and AI-driven interactions. A key factor in its effectiveness is the high quality of digitalization methods and realistic graphics, combined with the gamification module, which transformed the perception of the museum from being dull to a source of enjoyment and engagement. However, several challenges must be addressed to further improve the implementation of generative AI in the metaverse. These include optimizing the AI’s responsiveness to reduce lag and inconsistencies, enhancing the naturalness of the AI’s tone of voice, and addressing inaccuracies in AI-provided information. Additionally, hygiene concerns related to shared VR equipment must be resolved through proper cleaning protocols, especially for public museums or universities that plan to use shared virtual reality equipment.

This study has several limitations that should be considered in future research. First, the participant pool consisted primarily of university students, which may not fully represent the diverse usability needs and experiences of the broader population. Second, the study was conducted in a laboratory setting with high-performance PCs and advanced VR equipment, which may not reflect the typical hardware available to the general public. This controlled environment could influence both usability and system performance when compared to more modest or varied settings. Lastly, the duration of use was limited, potentially restricting participants’ ability to fully explore and interact with all system features. Future research should address these limitations by including a more diverse participant pool, testing the system in different physical environments and hardware contexts, and extending the usage period to evaluate sustained engagement and long-term usability (Chudasri et al., 2020; Nisi et al., 2020).

Future iterations of this system will incorporate AI-driven enhancements to address these challenges and improve the user experience. Specifically, dynamic non-player character tour guides, powered by conversational AI, will be developed to provide personalized and interactive experiences. These NPCs will offer context-specific guidance, address visitors’ questions through voice or text, and adapt their interactions based on user behavior and preferences. For example, NPCs could accompany visitors during their exploration, providing detailed insights into cultural artifacts and guiding them through gamified tasks. To further enhance immersion, advanced VR hardware, such as haptic gloves, will be integrated to replace traditional controllers, enabling more precise and natural interactions. Additionally, future research will expand the participant demographics beyond university students to include individuals from diverse age groups and cultural backgrounds, ensuring a broader and more inclusive evaluation of the system’s usability and effectiveness.