Selected finalist projects to be presented

June 13th – h: 10:00 – 11:15 CET 
(open conference room)

• Elton Vieira da Silva, Pedro Beça, Mónica Aresta
(DeCA., Universidade de Aveiro, Portugal)

Podcards-Edu: A Creative Pedagogical Tool for Podcast Creation in Education

Abstract:  The growing popularity of podcasts has opened new opportunities for integrating audio media into educational contexts, particularly through science communication podcasts. These formats encourage students to develop key competencies such as research, communication, collaboration, and creativity, aligning with active learning methodologies like Project-Based Learning and Learning by Teaching. Podcards-Edu is a pedagogical intervention designed to support podcast creation in diverse educational settings. Structured as a card game, Podcards-Edu guides learners through the five core stages of podcast production—Characters, Script, Recording, Editor, and Post-Production—using a visual and task-oriented approach. The tool includes a project-tracking board and a set of teacher-specific support cards offering implementation and assessment strategies. The intervention is adaptable across disciplines and educational levels, making it a replicable and inclusive strategy for promoting active learning through digital media.

• Alessia Nicoletta Marino
(Università di Siena, Italy)

MycorrhizAI

Abstract: MycorrhizAI is a project designed to create an informal smart learning space for high school and university students. Some studies show that young people primarily rely on the Internet and social networks for information, exposing themselves to the risk of fake news and opinion polarization. These mechanisms often arise due to algorithms used by digital platforms, which select and suggest content based almost exclusively on users’ preferences, thus reinforcing phenomena such as filter bubbles and echo chambers. The project aims to counter disinformation by promoting source literacy and transparency in information sharing. Moderators lead discussions on relevant topics, asking participants to properly inform themselves before contributing. Artificial intelligence is not used to reinforce individual biases, but rather to facilitate the encounter between different viewpoints. Just like mycorrhizae in nature foster communication between trees through underground networks, AI becomes a tool for connection and cooperation, capable of stimulating critical thinking, autonomous learning, peer collaboration, and moments of serendipity

• Mercedes Araya-Day
(Pontificia Universidad Católica de Chile)

LLM enhanced Small Smart Ecosystem Model

Abstract:  Pollution, climate change, and the childhood obesity epidemic are among the most urgent challenges of our time. Simultaneously, STEM education often fails to connect students with these pressing issues, remaining abstract and disengaged from lived experience. In response, we developed the LLM-SSEM (Low-cost Learning Model – Smart Small-scale Ecosystem Model), a hands-on, low-cost educational tool that simulates Earth’s and organisms’ metabolic processes inside a transparent plastic box. Equipped with CO₂ and temperature sensors, the system enables students to observe how different organisms—such as yeast and humans—consume energy and impact their environment. In this case study, we implemented the model in fourth-grade classrooms, where students engaged in active experimentation using their own breath and yeast fermentation. They recorded data, created graphs, and reflected through open-ended questions. We incorporated generative AI tools to provide feedback on learning: teachers could photograph a worksheet and receive real-time analysis of students’ responses. This combination of embodied experimentation with AI-driven evaluation defines the “smart” component of the model. Compared to a class that received teacher-led theoretical explanations, students in the experimental classes showed a significantly deeper understanding of core concepts such as metabolism, energy transformation, and the role of cells in sustaining life. Assessments were conducted using both manual analysis and Large Language Models (LLMs) with Chain-of-Thought reasoning, yielding consistent results across both methods, independent of any potential human biases. The LLM-SSEM is a promising tool for enabling an integrative exploration of science, nutrition, and climate change in a format that is accessible, portable, and engaging. So far, these results show that it reconnects science education with metabolism and the urgent realities shaping students’ futures.

• Wei-Jung Li, Cian-Yue Liao, Hui-Ni Chang, Yi-Jhen Lin
(National Cheng Kung University, Tainan, Taiwan)

A Virtual Reality-Based NIHSS Stroke Assessment Training System

Abstract:  Stroke assessment places high demands on speed and precision, but traditional lectures and video-based learning often rely on one-way information delivery, lacking interactivity and real-world applicability. To bridge this gap, we developed SS+, a Virtual Reality (VR) training system that helps medical students master the National Institute of Health Stroke Scale (NIHSS) through immersive, hands-on simulations. SS+ offers structured, repeatable training with diverse clinical scenarios, enabling learners to actively engage in clinical tasks and build confidence before working with real patients. As one student noted, “It’s easy to zone out when watching videos in class, but VR requires full engagement to receive meaningful feedback from virtual patients.” This aligns with previous research showing that hands-on practice enhances skill acquisition. In our comparative study, SS+ users demonstrated significantly higher motivation than those trained with videos. The system also received a “Good” rating on the System Usability Scale, supporting its acceptance and practical value in stroke education.