2026 ASEE Annual Conference & Exposition

Engineering technology education faces the ongoing challenge of equipping students with the skills to master complex concepts through authentic design experiences. A key opportunity lies in integrating artificial intelligence (AI) with simulation tools to foster design thinking, economic decision-making, and science-based knowledge, essential pillars for informed engineering decisions. This paper explores the implementation of AI, specifically large language models (LLMs), alongside simulation software to enhance learning by design in project-based settings.

Despite growing interest in emerging technologies, insufficient attention has been given to the synergy between LLMs and simulation-based design projects. The present study addresses this gap by grounding the approach in complex learning principles, applied within a first-year engineering technology course focused on energy efficiency. The intervention unfolded over two phases: an initial lecture-based session introducing energy principles, followed by assessments of conceptual understanding through energy knowledge questions and self-perception surveys on self-efficacy, trade-offs, and metacognition. This foundational knowledge prepared students for the subsequent design-based learning phase.

In the second phase, students received training on the Aladdin simulation software to model a house with solar panels affected by tree shading. Guided by design-based learning stages, they began with clarifying specifications, including budget constraints, net energy consumption goals, aesthetic and appeal considerations, costs, and environmental responsibilities. Students then used the simulation tool to generate design alternatives, select optimal configurations, prototype solutions, test performance, and iterate through redesigns. To augment this process, a conversational LLM was implemented as an AI agent, facilitating reflection and communication across design stages. Two AI variants were deployed: a scaffolded version providing structured prompts to elicit justifications, evidence from simulations, and trade-off analyses; and a non-scaffolded version enabling open-ended, human-like dialogue to finalize with the post-assessment.

Data from pre- and post-assessments, self-efficacy and metacognition surveys, and qualitative reflections demonstrated significant gains in students' conceptual grasp, decision-making confidence, and ability to articulate multifaceted design rationales. The integration of simulation for iterative prototyping and AI for reflective argumentation proved effective in deepening systems thinking and uncovering overlooked factors, such as sustainability trade-offs.

This work illustrates the feasibility of combining simulation tools with conversational LLMs in design-based learning, offering a scalable model for engineering technology education. By embedding AI-driven reflection into simulation workflows, the approach not only enhances technical proficiency but also promotes critical engagement with economic, social, and environmental contexts, paving the way for broader applications in reflective engineering practices.