● Led a team of four pre-final-year students to build a computer vision system for detecting mountings, verifying their position and orientation, and identifying missing ones on chassis frames, streamlining the quality control process.
● Developed a PoC by curating and augmenting a dataset of a sample chassis frame, trained a YOLOv8x model for mounting detection, and later replicated the results using YOLOv8n — achieving similar performance while reducing model size by 21×.
● Fixed camera coordinates to ensure consistent inspection input; developed set theory-based logic to detect missing mountings and their locations. 
● Explored and experimented with multiple approaches to verify orientation and position, yielding preliminary results; created a thorough report documenting methodologies and findings, and assisted in interviewing and onboarding the next batch of students, outlining future directions for the project.