Carlos Pereira Santos

This exploratory qualitative study examines how students’ reasoning becomes observable and develops within Socratic dialogues with Riffbot, a large language model (LLM)-based learning chatbot designed to guide students from descriptive and clarificatory questioning toward explanatory and reflective inquiry. The empirical dataset comprised 121 dialogue transcripts generated by 93 students across seven project-based courses at a university of applied sciences. All 121 transcripts were initially reviewed to examine participation, dialogue length, and interactional characteristics. Of these, 72 dialogues containing at least 10 turns were selected for detailed coding and reflexive thematic analysis (RTA), while 49 shorter dialogues were retained as contextual material. The analysis examined student responses in relation to successive chatbot prompts, using theoretically informed reasoning categories to trace interactional patterns within complete dialogues. Coded data were visualized in a chronological matrix to support the interpretation of reasoning trajectories. Descriptive reasoning predominated, while some dialogues showed emerging explanatory and reflective reasoning. Other trajectories remained primarily descriptive, involved repeated clarification, or moved recursively between reasoning categories, often including moments of uncertainty and redirection. These findings highlight the non-linear and interactionally situated nature of learning expressed during Socratic LLM–student exchanges, without establishing independent learning gains. Methodologically, the study demonstrates how systematic coding, chronological visualization, and reflexive thematic interpretation can be combined to make thinking-in-action more visible and analytically traceable. It contributes an empirical account of student reasoning within AI-mediated dialogue and a practical, digitally supported approach to investigating learning-related processes in higher education.

Current Digital Twins of the Ocean rely predominantly on two-dimensional geoportal interfaces that constrain the spatial comprehension of complex marine environments. This paper presents Immersive Ocean, a Virtual Twin platform developed under the EU-ILIAD Digital Twins of the Ocean initiative that procedurally transforms standardised European geospatial data (EMODnet, Copernicus Marine Service) into interactive three-dimensional maritime environments via Unreal Engine 5 (Epic Games, Cary, NC, USA). The platform supports both desktop and immersive virtual reality modes, enabling users to visualise and manipulate spatial scenarios involving offshore wind farms, shipping corridors, aquaculture installations, and species distributions. System performance testing confirmed stable rendering across PC and VR configurations (61 FPS and 42 FPS, respectively). A user evaluation with 31 participants across three hardware configurations revealed that core geovisualisation capabilities—ease of use, immersion, and procedural generation utility—remained robust, regardless of hardware, whilst feedback responsiveness (H = 5.99, p = 0.04995) and perceived realism (H = 6.31, p = 0.04258) were significantly affected. These findings inform deployment strategies for immersive geospatial tools: minimum specification systems preserve functional access, whilst recommended hardware enhances the perceptual fidelity critical for spatial decision support. This evaluation establishes usability among digitally proficient users; efficacy with domain stakeholders in authentic planning contexts remains for future work.

Digital Twins of the Ocean are increasingly adopted for Maritime Spatial Planning, yet most current implementations rely on 2D visualizations with limited interactive stakeholder engagement. The authors present Immersive Ocean, a VR-capable Virtual Twin platform developed under the EU-ILIAD initiative that procedurally generates 3D environments from standardized European geospatial data. The authors report system performance testing and a user experience study with 31 participants across three hardware configurations. Performance testing confirmed stable real-time rendering in both PC and VR. Comparative statistical analysis revealed that hardware configuration significantly affects feedback responsiveness and perceived realism, while core usability and procedural generation quality remained robust across configurations. These findings inform deployment strategies for environmental decision-support in Maritime Spatial Planning contexts. Immersive Ocean is currently undergoing piloting with real-world stakeholders, including applications in offshore wind planning scenarios, multi-use of space, and turbidity scenarios.

Digital Twins of the Ocean (DTOs) are increasingly used in Maritime Spatial Planning (MSP), yet most remain limited to 2D representations and offer minimal stakeholder interactivity. These limitations reduce their effectiveness in capturing complex socio-ecological-technical dynamics and supporting exploratory what-if scenario planning in a 3D or 4D ocean space. This paper presents Immersive Ocean, a novel Virtual Twin platform developed within EU-ILIAD DTO initiative. Built with game engine and VR technologies, it supports procedural 3D world generation and interactive exploration in both desktop and immersive VR modes. Systematic performance validation demonstrated stable frame rates across both PC and VR platforms. Initial user evaluations (n=22) report high usability and engagement but also suggest areas for improvement in UI clarity and ecological model representation. These initial findings position Immersive Ocean as a promising Virtual Twin solution for an immersive, interactive, and data-integrated approach to MSP and ocean governance. Immersive Ocean is now being piloted with stakeholders in real-world MSP scenarios, including offshore wind farm planning.

Our study introduces an open general-purpose platform for the embodiment of conversational AI systems. Conversational User-interface Based Embodiment (CUBE) is designed to streamline the integration of embodied solutions into text-based dialog managers, providing flexibility for customization depending on the specific use case and application. CUBE is responsible for naturally interacting with users by listening, observing, and responding to them. A detailed account of the design and implementation of the solution is provided, as well as a thorough examination of how it can be integrated by developers and AI dialogue manager integrators. Through interviews with developers, insight was gained into the advantages of such systems. Additionally, key areas that require further research were identified in the current challenges in achieving natural interaction between the user and the embodiments. CUBE bridges some of the gaps by providing controls to further develop natural non-verbal communication.

There has been significant progress in the graphical realism of digital humans in recent years. This work investigates the realistic portrayal of emotions beyond facial expressions by analysing how skin colour changes when expressing different emotional states. The study combines existing knowledge from old painters, photogrammetry data, thermal imaging, and skin colouration maps to create an artistic guideline to portray emotions realistically, resulting in the proposal of a set of colour maps representing the six basic emotions. By using skin colour changes to represent emotional states, the proposed colour maps offer an alternative workflow for portraying emotions. During the experiment of this research four of these proposed colour maps, which represent neutrality, anger, disgust, and happiness, were preferred over traditional alternatives in terms of realism perception and likeability. The findings have implications for the development of digital human technology, particularly in the creation of more realistic and expressive digital characters.

.Post-partum hemorrhaging is a medical emergency that occurs during childbirth and, in extreme cases, can be life-threatening. It is the number one cause of maternal mortality worldwide. High-quality training of medical staff can contribute to early diagnosis and work towards preventing escalation towards more serious cases. Healthcare education uses manikin-based simulators to train obstetricians for various childbirth scenarios before training on real patients. However, these medical simulators lack certain key features portraying important symptoms and are incapable of communicating with the trainees. The authors present a digital embodiment agent that can improve the current state of the art by providing a specification of the requirements as well as an extensive design and development approach. This digital embodiment allows educators to respond and role-play as the patient in real time and can easily be integrated with existing training procedures. This research was performed in collaboration with medical experts, making a new contribution to medical training by bringing digital humans and the representation of affective interfaces to the field of healthcare.

The recent success of Machine Learning encouraged research using artificial neural networks (NNs) in computer graphics. A good example is the bidirectional texture function (BTF), a data-driven representation of surface materials that can encapsulate complex behaviors that would otherwise be too expensive to calculate for real-time applications, such as self-shadowing and interreflections. We propose two changes to the state-of-the-art using neural networks for BTFs, specifically NeuMIP. These changes, suggested by recent work in neural scene representation and rendering, aim to improve baseline quality, memory footprint, and performance. We conduct an ablation study to evaluate the impact of each change. We test both synthetic and real data, and provide a working implementation within the Mitsuba 2 rendering framework. Our results show that our method outperforms the baseline in all these metrics and that neural BTF is part of the broader field of neural scene representation. Project website: https://traverse-research.github.io/NeuBTF/.

Steenbeek, J., G. Romagnoni, J. W. Bentley, J. J. Heymans, N. Serpetti, M. Gonçalves, C. Santos, H. Warmelink, I. Mayer, X. Keijser, R. Fairgrieve, and L. Abspoel. 2020. Combining ecosystem modeling with serious gaming in support of transboundary maritime spatial planning. Ecology and Society 25(2):21. https://doi.org/10.5751/ES-11580-250221