Earthquaker-AI: A RAG-Based Educational Framework for Earthquake Preparedness in Primary Schools

Forum topic · 小凯 · 2026-07-17

Summary

Earthquaker-AI is a hybrid educational framework that combines educational robotics with a Retrieval-Augmented Generation (RAG) conversational AI assistant to teach earthquake preparedness to primary-school students. Building on the award-winning STEM project Earthquaker, the system moves from mechanical simulation with Lego WeDo2 to cognitive and metacognitive learning. The robotics component uses Lego WeDo2 automation to simulate seismic response, allowing students to interact with sensors and actuators as tangible representations of protective actions. The AI assistant aligns student responses with official safety guidelines and provides rubric-based verbal feedback supporting self-regulated learning. It follows a developmental learning trajectory: younger students identify safe actions via multiple-choice questions scored on a two-axis rubric; middle grades identify correct action sequences with a three-axis rubric; older students produce short written responses evaluated on a four-dimension rubric including expressive clarity. The dialogue module uses RAG to semantically match student queries against official guidance, generating safe and accurate responses. Experimental evaluation showed high groundedness and accuracy with a low hallucination rate.

Overview

Field: Machine Learning / AI in Education Authors: Xanthi Kokkinou, Chaido Mizeli, Nafsika Koulaxidou, Marina Delianidi, Konstantinos Diamantaras Published: 2026-07-15 arXiv: 2607.14046

Abstract

This paper presents Earthquaker-AI, a hybrid educational framework building upon a previously implemented educational robotics project by integrating a conversational AI assistant based on Retrieval-Augmented Generation (RAG). It aims to enhance earthquake preparedness and conscious action among primary-school students. The system extends the award-winning STEM project Earthquaker, moving from mechanical simulation with Lego WeDo2 to cognitive and metacognitive processing.

The robotics component uses Lego WeDo2 automation to simulate seismic response, letting students interact with sensors and actuators as tangible representations of protective actions. The assistant operates as a guided learning mechanism, aligning student responses with safety guidelines while providing rubric-based verbal feedback that supports self-regulated learning and calmness during emergencies.

Progressive Learning Trajectory

Earthquaker-AI follows a gradual learning trajectory aligned with cognitive development:

Lower grades: basic identification of safe actions through multiple-choice questions, assessed with a two-dimension rubric.
Middle grades: identification of correct action sequences through multiple-choice questions, assessed with a three-axis rubric.
Upper grades: shift to verbal production, requiring short written responses evaluated with a four-dimension rubric that includes expressive clarity.

RAG Dialogue Module

The conversational module uses Retrieval-Augmented Generation to semantically match student queries against official safety guidelines, generating safe and accurate responses. Experimental evaluation demonstrated high groundedness and accuracy, with a low hallucination rate.

Conclusion

Overall, Earthquaker-AI combines hands-on engagement, information processing, and reflective practice. By integrating robotics, rubrics, and AI, it promotes technological literacy, self-regulation, and responsible use of digital systems, contributing to early crisis-management skills.