EduBot: A Low-Cost Multilingual AI Educational Robot for Inclusive and Scalable Learning

摘要

This paper presents EduBot, a novel, low-cost, AIpowered educational robot designed to foster inclusive and scalable learning for Bangla and English-speaking communities. EduBot addresses critical limitations in existing educational robotics-high cost, restricted language support, and lack of contextual adaptability-by integrating multilingual interaction, Optical Character Recognition (OCR), speech recognition, face detection, and deep learning-based natural language processing within a compact, affordable system. Built using open-source software and economical hardware, EduBot delivers real-time interactive learning via a mobile application. Capabilities include voice-enabled Q&A, book scanning, storytelling, math problemsolving, and expressive gestures. Experimental results demonstrate EduBot’s superior performance: achieving OCR accuracies of 95.2% for Bangla and 97.6% for English, a response time of 150 ms, and significantly higher user satisfaction compared to existing solutions. Its modular architecture enables deployment across traditional, hybrid, and remote learning environments. Furthermore, its culturally relevant design ensures accessibility in resource-constrained settings. This work advances educational technology by offering a practical, scalable platform for personalized, AI-driven learning. EduBot establishes a new benchmark for low-cost, multilingual educational robots and presents a promising model for future integration into global learning ecosystems.