Conference Agenda

The transition toward circular economy principles and Industry 5.0 para-digms is driving the need for effective and human-centered training solutions for industrial disassembly operations. Virtual reality technologies offer im-mersive and risk-free learning environments for operator training; however, structured methodologies for quantitatively validating these systems are still limited. This work presents a novel validation approach for manual disas-sembly tasks of mechatronic products using VR-based industrial training ap-plications. The approach consists of a system stabilization phase followed by a large-scale validation phase based on objective performance indicators. Experimental activities were conducted within the X-DE-MAN project on household appliance disassembly scenarios involving sixty-four participants. Comparative tests between immersive VR training and conventional paper-based instruction were carried out to evaluate learning effectiveness. The ob-tained results demonstrate that the proposed training approach reduces error rates, increases proficiency scores, and provides more consistent perfor-mance among users with heterogeneous backgrounds and prior experience. Furthermore, the VR-generated proficiency assessment proved to be a relia-ble indicator of actual operator capabilities.

Physical laboratories in industrial and academic contexts are frequently geo-graphically dispersed, making simultaneous collaborative operation across sites technically complex and resource intensive. This paper presents a serv-er-authoritative distributed infrastructure that connects independent virtual laboratory environments within a single shared real-time session, enabling geographically distributed users to observe and interact with simulated indus-trial processes simultaneously, without per-site network configuration. The architecture is built on Unity Netcode for GameObjects and Unity Relay, and is designed so that each laboratory is an autonomous, independently opera-ble module that can be added to the shared session without modifying exist-ing components. To lower the technical barrier for adoption, the infrastruc-ture is accompanied by a Unity Editor plugin that automates the creation and pre-configuration of all required networking components, leaving developers only the task of installing four prerequisite packages and supplying session-specific parameters. The approach is validated through a scenario of three connected laboratory environments simulating sequential industrial process-es. Runtime evaluation across six network configurations yields a mean frame rate exceeding 350 FPS and a global mean command latency of 2.52 ms, confirming suitability for real-time distributed collaborative use.

Digital Twins and Extended Reality (XR) technologies are increasingly adopted to support industrial design validation, process planning and human-centred in-teraction. This study presents an open-source workflow for a robotic milling sys-tem using Blender as a unified environment for workspace reconstruction, 6-DOF robot rigging, path-based motion simulation, surface-model preparation and immersive inspection. The virtual representation supported the evaluation of al-ternative machine positions, robot reach and spatial clearances before physical in-stallation. A preliminary user study involving 23 participants assessed perceived usability and user experience through the System Usability Scale (SUS), User Experience Questionnaire (UEQ) and qualitative feedback. The SUS score of 73.8 and positive UEQ results indicate favourable user perception, particularly regarding immersion and interaction quality. These findings concern usability and do not constitute a quantitative validation of simulation accuracy or engineering performance. The results suggest that the proposed workflow may support pre-liminary layout assessment and immersive review in industrial robotic applica-tions.

Mountain emergencies often require companions or other non-expert people to perform initial safety and first-aid actions before professional rescuers arrive. However, traditional training offers limited opportunities to practice these procedures in realistic, safe, and repeatable outdoor emergency scenarios. This paper presents an immersive Virtual Reality application for mountain emergency training, aimed at non-expert users, that simulates a scenario in which the user assists an unconscious companion during a mountain excursion. The experience includes a guided training phase, and an unguided test phase with more challenging environmental conditions. A pilot study with 25 participants was conducted to assess the feasibility of the workflow and collect preliminary data on procedural performance, perceived workload, and emotional engagement. Results showed that participants generally completed the procedure after training, while the main difficulties concerned decision-making and time efficiency. Self-report measures indicated that the test phase was perceived as more demanding, less controllable, and more stressful than the guided training phase. Although exploratory, the study suggests that immersive VR can support safe and repeatable training of mountain emergency procedures for non-expert users.

Traditional astronomical heritage preservation often relies on static exhibits, leaving visitors to attend the tour's narrative passively. This paper introduces a Virtual Reality (VR) Digital Model (DM) of the historic Ruths telescope. Designed to overcome the geographical and meteorological limitations of physical observatories, it empowers visitors to experiment with a simulation-driven virtual replica in the same way as the real counterpart.

By combining an immersive virtual environment with a simplified backend and physical prototype, we transform a complex mechanical instrument into an interactive educational resource. This approach shifts the focus of digital cultural heritage preservation from passive observation to active engagement, enabling audiences to safely operate and understand the functioning of complex historical machinery while preserving the integrity of the original asset.

Virtual Reality (VR) offers new opportunities for safety training by enabling workers to practice hazardous tasks in controlled and interactive environments. However, the deployment of VR solutions within real industrial training programs raises practical challenges related to usability, acceptance, integration, and user diversity. This paper presents the real-world deployment of a VR-based safety training application for forklift inspection and hazard identification. The application was integrated into regularly scheduled workplace safety training activities in two industrial companies and tested with 36 participants from heterogeneous professional backgrounds, including operators, safety personnel, supervisors, and managerial staff. The study combined usability assessment, performance measures, movement tracking, eye-tracking data, interaction logs, and qualitative feedback. Results show that the application achieved good levels of perceived usability and that participants rapidly developed more structured exploration strategies between consecutive exercises. Movement and gaze data further highlighted the potential of VR not only as a training tool, but also as a means to analyze user behavior and support targeted feedback. At the same time, qualitative feedback revealed challenges related to initial familiarization, physical comfort, interaction realism, and organizational adoption. The paper discusses practical implications for the design and deployment of VR safety training in industrial contexts.