Abstract:
This discussion paper investigates short Manim based animations as reusable learning objects for undergraduate service mathematics and addresses four questions on pedagogical value, production effort, the role of generative AI, and design principles. A part of our contribution is a reproducible workflow, including concrete engineering choices for camera orientation, label management, and render time control through a quality flag. The paper further outlines a git centered development process with AI assistants and records practical guidance on Manim Community Edition usage and agentic code generation. A small deployment was conducted in a small Analysis course within a civil engineering dual program in the summer semester 2025 at the IU Internationale Hochschule. Likert scale results indicate positive perceptions of relevance and clarity, and YouTube analytics show peaks in viewing near release dates during the lecture period from April 2025 to June 2025 with no spike during the exam phase. This pattern suggests use for in term understanding rather than last minute exam preparation. Qualitative comments cite improved visualization of abstract content and clearer answers to why certain topics matter, with a request for tighter links to worked examples. The paper synthesizes design considerations for guided and self-guided consumption, timing and narration, and constraints imposed by small displays, and motivates reuse and open sharing of sourcecode-based animations. Limitations include a single site context, small sample, and the absence of objective learning outcome measures.
