A Bayesian framework for multilead SMD post-placement quality inspection

Area of Science:

• Manufacturing Engineering
• Quality Control Systems
• Robotics and Automation

Background:

• Surface Mount Technology (SMT) is critical in modern electronics manufacturing.
• Accurate placement of Surface Mount Devices (SMDs) on Printed Circuit Boards (PCBs) is essential for device functionality and reliability.
• Current inspection methods may lack the precision needed for real-time, post-placement quality assessment.

Purpose of the Study:

• To propose a novel framework for inspecting the placement quality of SMDs immediately after placement on PCBs.
• To develop an indirect measurement technique for lead displacement relative to ideal pad positions.
• To enhance the accuracy of lead displacement computation through a combined classification/estimation process.

Main Methods:

• Indirect measurement of lead displacement using image data area.
• Classification process to infer displacement from raw image data of lead regions.
• Combined classification/estimation process to improve accuracy of lead displacement computation.
• Derivation of geometric relations connecting lead shifts to rigid body component displacement.

Main Results:

• A novel framework for SMD placement quality inspection was developed.
• The method accurately infers individual lead displacement from image data.
• Geometric relations were derived to refine lead shift measurements, crucial for quality assessment.
• Experimental results demonstrated the algorithm's potential for accurate quality control.

Conclusions:

• The proposed framework offers a promising solution for real-time inspection of SMD placement quality.
• The developed indirect measurement and refined calculation methods significantly improve accuracy in assessing lead displacement.
• This approach contributes to enhanced quality control in PCB manufacturing.