IEEE AIxVR 2025 is the premier international conference dedicated to the convergence of artificial intelligence (AI) and extended reality (XR). The event serves as a platform for researchers, industry leaders, and developers to share cutting-edge advancements in AI-powered XR applications. AIxVR 2025 showcased groundbreaking research, interactive demonstrations, and thought-provoking discussions, making it an essential gathering for those shaping the future of immersive technologies.
This year’s edition took place in Lisbon and brought together researchers, practitioners, and innovators from around the world. The conference provided a unique opportunity to present research, engage with industry experts, and explore the evolving landscape of AI-powered XR applications.
Presentation: User-Centric Evaluation Methods for Digital Twin Applications in Extended Reality
One of the highlights of my participation was presenting my short paper, User-Centric Evaluation Methods for Digital Twin Applications in Extended Reality. This research, conducted as part of the DIDYMOS-XR project, focused on establishing user-centric evaluation frameworks for XR-enhanced Digital Twins (DTs). Given the increasing adoption of DTs across industries — from smart cities to industrial maintenance — the need for systematic assessment methods is crucial.
During my presentation, I received intriguing questions from the audience, particularly about the relationship between Presence and User Experience (UX) in DT applications. The discussion revolved around how an enhanced sense of presence in XR environments affects usability, engagement, and decision-making, especially in industrial applications where precision and real-time interactions are critical. Another key point of interest was the industrial use case, where DTs are employed to optimize real-world operations by integrating real-time data with immersive simulations.
DIDYMOS-XR Special Session: Creating and Updating Digital Twins for Enabling XR Applications
Another highlight of AIxVR 2025 was the special session organised by the DIDYMOS-XR project, Creating and Updating Digital Twins for Enabling XR Applications. This session hosted five oral presentations, bringing together experts from academia and industry to explore innovative approaches to DT development and application. One paper presented the collaborative approach to reconstruction used by DIDYMOS-XR, which involves fusing data from multiple cameras mounted on a vehicle, LiDAR data, and static cameras. Other papers addressed the problem of accurate localization from various angles. One approach used only GPS and odometry data, while another combines visual and LiDAR data for cooperative localization of multiple agents. A paper for vision-based localization combined traditional feature matching with learned semantic segmentation and depth estimation, while another one proposed a localization approach based on matching views reconstructed from a NeRF (Neural Radiance Field); a method for high-quality 3D construction of images. You can read more detail on the kinds of technical work undertaken by DIDYMOS-XR on our publications page.
A particularly engaging part of the session was the feedback on AI-based localization algorithms. Participants discussed challenges in real-time localization for DTs and how AI-driven solutions can enhance accuracy and reliability. Additionally, the session sparked a productive conversation about different use cases and validation methods, emphasizing the need for standardized evaluation techniques to ensure the practical applicability of DT solutions across various industries.
Key Takeaways and Future Directions
AIxVR 2025 provided invaluable insights on the intersection of AI, XR, and Digital Twins. The discussions reinforced the importance of user-centric design and evaluation in the development of XR-enhanced DT applications. As these technologies continue to evolve, ensuring their usability, efficiency, and adaptability to real-world scenarios will be crucial.
The feedback and discussions at the conference will undoubtedly shape the next steps in our research within the DIDYMOS-XR project. The focus on user experience, presence, and AI-enhanced localization will guide future studies, ensuring that XR and DT solutions meet the needs of their end-users effectively.
A huge thanks to the organizers of AIxVR 2025, the participants in my session, and everyone who engaged in thought-provoking discussions. We are looking forward to seeing how these insights translate into advancements in Digital Twin applications and beyond!
Author: Francesco Vona, Hochschule Hamm-Lipstadt, with additional material by Werner Bailer, Joanneum Research
