VokaSight
My Role
Fullstack Developer
Automated hardware and soldering inspection mobile application for vocational labs. Detects bridging short circuits, solder voids, and component misalignment via smartphone camera.
Context
Hackathon
CHALLENGE
Vocational instructors face severe time bottlenecks manually assessing dozens of student electronic assemblies. Evaluating soldering precision and component orientation is subjective, slow, and risks costly component blowouts when undetected short circuits are powered on.
APPROACH
Developed VokaSight, an end-to-end mobile solution combining freehand smartphone image normalization (OpenCV perspective transformation & CLAHE contrast tuning) with real-time YOLOv8-Nano defect inference to identify solder bridges, voids, and inverted IC/diode polarities.
OUTCOME
Delivered a high-accuracy grading companion that pre-validates student PCB assemblies in seconds, surfaces clear visual defect bounding boxes, and generates objective scores before physical circuit testing.