Automated PCB Fault Detection System Using Embedded Computer Vision on Conveyor Lines

Objective: To design and implement an embedded computer vision system for real-time detection of faults in printed circuit boards on automated conveyor lines.

In Indian electronics manufacturing plants, manual inspection of printed circuit boards (PCBs) for defects such as missing components, soldering faults, and misalignments is labor-intensive, error-prone, and slows down production lines. Small- and medium-scale industries often lack affordable, automated solutions, leading to increased rejection rates and customer returns.

This project proposes an embedded computer vision system using a Raspberry Pi or NVIDIA Jetson Nano, integrated with high-resolution cameras and OpenCV-based algorithms, to detect visible PCB faults in real-time as boards move on a conveyor line. The prototype will be built and demonstrated using a custom conveyor setup, lighting control, and a dataset of common PCB faults relevant to Indian industry.

Key deliverables include: a fabricated conveyor testbed, embedded vision hardware, trained fault-detection model, real-time defect classification and alerting interface, and comprehensive validation against industry-standard test PCBs. The system will be benchmarked for accuracy, speed, and practical deployment feasibility.

This solution enables affordable, scalable quality control for Indian PCB manufacturers and EMS providers, reducing manual inspection errors and improving throughput. The approach can be adapted for various board types and scaled for larger production environments, supporting the 'Make in India' initiative.