Development of interactive media digital applications through a deep learning approach in the automotive concentration

Abstract

This research aims to develop interactive learning media based on digital applications that can improve students' conceptual understanding and skills in the automotive concentration through the application of a deep learning approach. The development of this media is motivated by the need for learning innovations that integrate digital technology with deep learning strategies so that students not only memorize procedures but also comprehensively understand the working principles of automotive systems. The research method used is Research and Development (R&D) with the Borg & Gall model which includes the following stages: (1) needs analysis, (2) product design, (3) interactive media development, (4) validation by media experts and material experts, (5) limited trials, and (6) revisions and field trials. The research subjects came from students of the Automotive Engineering Education study program at Muhammadiyah University of Purworejo. Data collection used surveys, in-depth interviews, and Focus Group Discussions (FGD) in designing learning media. The results of the study indicate that the interactive media developed using the deep learning platform showed that the indicator with the highest score was ease of access (4.50), followed by industry relevance (4.40) and feedback (4.40), which both ranked second. Next, material completeness (4.30) and technological innovation (4.30) ranked next, followed by interactive displays (4.20). Automatic evaluation indicators (4.10) and simulation quality (4.10) ranked next with the same score. The lowest scores were for learning flexibility (4.00) and online collaboration (4.00), indicating a need for further development in both aspects. Overall, this ranking demonstrates the platform's superior accessibility and industry relevance, although improvements in flexibility and online collaboration are still needed. Therefore, this interactive media, based on digital applications using the Deep Learning approach, is effective as an innovative learning tool in the automotive sector.