Development of an AI-powered assistive technology-based learning model to improve learning accessibility for vocational students with disabilities

This study aims to develop an AI-Powered Assistive Technologies–based learning model to enhance learning accessibility for vocational students with disabilities. Using a Research and Development (R&D) approach and adopting the ADDIE model up to the development stage, this research produced a prototype of an adaptive and inclusive learning model equipped with intelligent assistive features, including text-to-speech, speech-to-text, automatic captioning, visual adaptation, and AI-driven navigation support. The needs analysis revealed significant accessibility barriers in vocational learning environments, particularly the limited availability of adaptive materials and insufficient assistive technology integration for students with sensory, cognitive, and motor impairments. The design and development phases resulted in a multimodal, AI-assisted learning framework aligned with principles of inclusive pedagogy. Expert validation involving specialists in vocational content, assistive technology, inclusive pedagogy, and language yielded an overall feasibility rating of 96.5%, categorised as highly feasible. In addition, user response testing involving 20 vocational students with disabilities demonstrated a positive acceptance rate of 99.4%, indicating that the AI-powered assistive features effectively enhanced engagement, comprehension, and independent learning. The findings affirm that the proposed model offers strong potential to support inclusive vocational education by addressing accessibility gaps and improving the learning experiences of students with disabilities. Although the study has not yet progressed to the implementation phase, the developed prototype provides a robust foundation for further empirical testing and large-scale application. Continued refinement is recommended to strengthen the ethical, pedagogical, and technological dimensions of AI integration in accessible learning environments.