Decision Framework for Generative AI Adoption in Engineering Education: A Practical Guide from the Global South Perspective

This article introduces a replicable and verifiable research framework for generative AI adoption, focusing on the field of engineering education, with special attention to implementation challenges and solutions in the context of the Global South and resource inequality. The framework aims to provide decision support for education managers, ensuring the fair and effective application of AI tools in resource-constrained environments.

Generative AI has profoundly transformed education, but existing research mostly focuses on resource-abundant developed countries, neglecting the infrastructure constraints, digital divide, and diverse educational backgrounds in the Global South. As a core area for training technical talents, engineering education requires a systematic framework to guide AI adoption decisions, in order to address the issue of fair benefit in resource-constrained environments.

This framework adheres to the principles of replicability, context sensitivity, decision orientation, and ethical considerations, and consists of three main components: 1. Environmental assessment dimensions (technical infrastructure, human resources, institutional support, cultural background); 2. Adoption phase model (Cognition → Exploration → Adaptation → Integration → Innovation); 3. Impact factor analysis (driving factors, hindering factors, moderating variables).

For resource-constrained environments, the framework proposes four strategies: 1. Offline-first design (supporting offline/low-bandwidth tools and alternative solutions); 2. Gradual introduction (starting with pilot projects in individual courses and then expanding); 3. Community-driven support (teacher-student mutual assistance communities, using peer learning to make up for insufficient technical support); 4. Cost-benefit analysis (assessing tool subscriptions, training costs, and teaching benefits).

Based on the framework, education managers can take the following steps: 1. Baseline survey (assessing technical readiness and the attitudes of teachers and students); 2. Pilot design (selecting 1-2 courses for pilot projects and setting evaluation indicators); 3. Data collection (gathering usage data, learning outcomes, and feedback); 4. Iterative optimization (adjusting strategies based on pilot results); 5. Scaled promotion (developing a phased expansion plan); 6. Continuous evaluation (regularly monitoring effects and updating strategies).

This framework fills the gap in generative AI education research and supports researchers from the Global South to participate in relevant discussions. In the future, it will be updated with technological evolution to incorporate new capabilities and scenarios, with the goal of establishing a global knowledge base that collects best practices under different resource conditions, providing empirical support for the digital transformation of engineering education.