Advancing Engineering-Driven STEM Pedagogy with AI

Abstract

Engineering-driven STEM pedagogy promotes integrative problem-solving through the engineering design process ( EDP ), but its implementation is constrained by curriculum integration, assessment alignment, and instructional workload. Recent advances in generative AI (GenAI) create possibilities for design-based learning through ideation, modeling, feedback, and documentation. However, AI ’s educational value depends on phase-specific integration within human-centered pedagogical frameworks, especially in engineering contexts where precision, reliability, and contextual validity are critical. This record-level systematic synthesis examines 50 Scopus-indexed records at the intersection of engineering education, higher education, and teacher AI literacy, with attention to AI integration and human– AI collaboration. Findings suggest that AI is most defensible when it augments human agency within EDP -aligned workflows. Concerns about reliability, contextual adequacy, equity, and integrity highlight the need for verification routines and governance. Teacher AI literacy emerges as a bridging construct connecting technology and pedagogy. Future research should test collaboration scaffolds, AI -mediated assessment, and engineering-specific AI literacy development.