Bridging Digital Literacy Gaps Through Learning-By-Doing: An IOT Project-Based Curriculum for AI Literacy in Vocational Education

Abstract

This article examines curriculum innovation through the implementation of the institutional compulsory course, "Introduction to Data Literacy and Artificial Intelligence," at the Institut Teknologi Sains Bandung (ITSB) during the even semester of the 2025/2026 academic year. The palm oil processing industry is rapidly transitioning toward automation and Industry 4.0 practices, demanding graduates equipped with data-driven practical skills (future skills) rather than mere mechanical competence. However, the implementation of this course faces significant demographic and pedagogical challenges: most students, primarily originating from outside Java, enter the program with limited prior experience in computer operations, yet rely heavily on generative AI platforms to complete assignments through uncritical copy-pasting—a pattern that erodes genuine conceptual understanding. To address this dual gap, this study introduces and evaluates a learning-by-doing methodology grounded in constructionism and experiential learning theories. Instead of theoretical assignments easily solved by AI, students are directly confronted with authentic technical problems (directly attach the problem): assembling simple electronic circuits based on the ESP32 microcontroller and sensors, and subsequently building fully functional mobile sensor-reading applications using the block-based drag-and-drop platform, MIT App Inventor. This approach is designed to position AI as a proportional assistant (co-pilot) that explains logic rather than acting as a substitute for cognitive processes, transforming students from passive consumers of AI text into active technology designers. Utilizing a descriptive qualitative case study design across three learning stages, this research reveals that visual block-based programming successfully bridges the coding capability gap for non-informatics students, enabling them to comprehensively understand the data flow—from sensor acquisition to data logging and cloud visualization—which is essential for adapting to an increasingly automated palm oil industrial environment. The article concludes with pedagogical implications, research limitations, and recommendations for expanding this approach to other vocational courses.