Development of an AI-based Spanish-Japanese Voice Translator for Engineering Education

Abstract

Linguistic diversity represents a significant challenge in engineering education, particularly in contexts that require international collaboration and access to multilingual technical content. Although commercial machine translation tools are widely available, few solutions are specifically designed to support pedagogical objectives in engineering training environments. This study presents the design and implementation of an artificial intelligence (AI)-based Spanish–Japanese voice translator with a modular architecture tailored for educational use. The system integrates speech recognition, optical character recognition, and machine translation components and was evaluated under controlled conditions using typed text, handwritten input, and voice data. Experimental results show an overall accuracy above 94%, with particularly strong performance in handwritten text recognition (98%) and Spanish audio transcription (96%). Beyond translation functionality, the proposed system supports project-based learning (PBL) by enabling students to interact with and modify AI modules, fostering competencies in natural language processing and intelligent systems design. The findings suggest that AI-driven multilingual tools can enhance digital inclusion and intercultural communication in engineering education contexts.