Engineering students face difficulties with abstract ideas and bridging theory with practice. This study examines how to design virtual reality environments sustaining engagement beyond initial novelty. Researchers developed a multi-modal system integrating CAVE technology, MATLAB simulations, and AI motion tracking in project-based learning. Twenty mechanical engineering undergraduates participated in a four-week “Biomechanics of a Pitch” project within a Strength of Materials course. Using mixed-methods analysis with validated instruments (α = 0.92–0.93) and focus groups, researchers measured presence, representation fidelity, and situational interest. Results showed CAVE excelled in spatial presence (α = 0.93) and immersive understanding (M = 4.2, SD = 0.8), MATLAB supported analytical thinking, while AI tracking enabled personalization. Student interest evolved from curiosity through sustained engagement to post-experience motivation. Strategic transitions between modalities maintained engagement beyond novelty. The authors present a framework integrating interest development theory with multi-modal VR design principles and provide implementation guidelines for engineering educators.