Zhentao Bao¹ and Rodrigo Onate², ¹USA, ²California State Polytechnic University, USA

This research addresses the barriers children face in accessing quality tennis instruction, particularly those with special needs who require personalized learning approaches. Traditional programs often lack individualized teaching methods and accessibility, with approximately 60% of children dropping out of organized sports by age 13 due to cost, limited access, and engagement challenges. To overcome these issues, the proposed AI-powered tennis training application leverages OpenAI's ChatGPT API, Firebase cloud infrastructure, and progressive level systems to deliver personalized, narrative-driven training experiences [1]. Its three core systems include AI content generation for individualized exercise descriptions, Firebase authentication for secure user management, and real-time progress tracking with visual feedback [2]. Key design challenges were resolved through caching strategies, fallback content mechanisms, and autism-specific personalization features. User testing with five participants yielded an average satisfaction score of 4.32/5.0, with particularly high ratings for age-appropriate content (4.8/5.0) and progression clarity (4.6/5.0). By eliminating cost barriers and offering therapeutic benefits through individualized AI-generated content, this solution makes tennis instruction more accessible while fostering sustainable engagement through imaginative, personalized adventure narratives.

AI-powered training, Personalized learning, Tennis accessibility, Autism-friendly design