Students in STEM disciplines frequently encounter challenges when asked to visualize complex three-dimensional (3D) structures from static two-dimensional (2D) textbook images. This translation requires significant spatial reasoning, which often becomes a barrier to mastering key concepts in engineering and other fields. Traditional instructional tools, while valuable, provide limited interactivity and do not fully support diverse learning styles.
To address this gap, we developed 3D STEMulate (a combination of '3D + stimulate' and 'STEM'), a web-based platform that transforms textbook diagrams and textual descriptions into interactive 3D images. Using a simple interface, students and instructors can upload any picture or enter the name of a scientific concept. An embedded AI model interprets this input and automatically generates executable code to render a manipulable 3D image. Users can then pan, rotate, zoom, and explore the visualization from multiple angles, thereby enhancing spatial comprehension.
Beyond screen-based interaction, the system can project generated 3D images onto a custom holographic display, offering a true 3D experience without the need for special glasses. This multimodal approach bridges the gap between traditional static 2D course content and immersive learning, creating new opportunities for engagement and deeper understanding. Importantly, 3D STEMulate is not limited to undergraduate or graduate STEM instruction within engineering; its applicability extends across multiple educational levels and disciplines.
The project contributes to engineering education by providing a scalable, accessible, and low-cost solution that integrates cutting-edge AI visualization with classroom learning. To evaluate its impact, 3D STEMulate was initially tested by the developers, confirming its usefulness. Further testing is currently being piloted in an electromagnetism course with three groups of students. Two groups consist of students with no prior knowledge of the tested concept: one taught using traditional methods, and the other with a blended approach that incorporates the interactive 3D STEMulate platform. A third group consists of former students who had previously completed the course and are now revisiting the material with 3D STEMulate. Early results indicate that students who engaged with the interactive tool demonstrate stronger comprehension of the concepts compared to those taught with traditional methods alone. We are continuing to test 3D STEMulate, collect feedback, and evaluate user experiences, with the goal of refining the tool to make it more effective and better tailored to the needs of students and instructors.
The full paper will be available to logged in and registered conference attendees once the conference starts on June 21, 2026, and to all visitors after the conference ends on June 24, 2026