A Pilot Study: Engineering Students Use Generative AI to Support the Development of Playful Educational Technology

This pilot study investigates how third-semester engineering students used generative AI tools to support a 20-week project developing playful educational technology. Drawing on qualitative design-based research, the study analyzed written reflections and classroom data from 36 students who worked with tools such as ChatGPT, Copilot, StableDiffusionXL, and ElevenLabs. The research centers on Danish engineering education, and its findings are presented as exploratory rather than generalizable, given its small scale and the absence of quantitative outcome measures.

Students most commonly turned to generative AI for brainstorming, debugging, graphics creation, proofreading, and other time-saving tasks, typically treating its output as a teammate whose suggestions required validation. Their output was rarely used without adaptation. Participants consistently reported that the tools produced hallucinations requiring fact-checking, and many found that crafting effective prompts consumed considerable time. Students also avoided applications where they perceived ethical concerns.

The authors conclude that generative AI can be productively integrated into playful educational technology development in contexts where it is efficient, reliable, and ethically acceptable. They argue that structured reflection requirements can foster AI literacy among engineering students. The study also calls for continued research and clearer academic-integrity guidance to govern classroom use of these tools.