Year: 2026
Type: conference-paper
Abstract: In the evolving landscape of online STEM education, ensuring student engagement and success in remote laboratory environments remains a challenge. The OELAssist project integrates AI-powered anomaly detection within the Open Engineering Lab (OEL) to provide timely and adaptive learning support. By applying isolation forest models, OELAssist analyses real-time experimental interactions to identify atypical student behaviours indicative of struggles during practical activities. By detecting deviations from expected experimental workflows, the system enables proactive interventions, reducing attainment gaps and enhancing retention. Initial findings suggest that students exhibiting multiple anomalous interactions (such as prolonged motor activation or inconsistent parameter adjustments) may benefit from additional guidance. This session will offer an interactive exploration of OELAssist, demonstrating how AI models analyse student engagement in online laboratories. We will demonstrate the experiment and the collection of real-time data. Then, through a hands-on exploration of the data and analysis, attendees will experience how AI flags anomalies and recommends interventions. The discussion will highlight the broader impact of AI-driven learning support and its potential applications across STEM disciplines.
Authors: Kbaier, Dhouha, Mason, Andrew
Language: English
Subfield: Computer Science Applications
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7009575515
FWCI:
Cites:
Cited by:
Related to: