PCB Inspection Training Boards for AI Defect Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional AOI devices for PCB assembly lack intelligence, leading to either missed defects with loose judgement criteria or excessive false positives with strict criteria, necessitating manual inspection and increasing human error.

Innovation Solution

Utilize visual inspection training boards with known defects to train an AI algorithm in AOI systems, including physical and virtual boards, to accurately identify defects like incorrect component placement, solder misregistrations, and solder defects.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If judgment criteria are tightened to detect all defects, then detection capability is improved, but false positive rate increases requiring manual inspection

Engineering Contradiction:

Improvedefect detection capabilityVSAvoidfalse positive rate

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The system performs preliminary training using visual inspection training boards with known defects before actual production inspection. This preliminary action allows the AI algorithm to learn correct defect identification patterns, reducing false positives during actual use while maintaining high detection accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses feedback from comparing AI predictions against known defect labels in training data to continuously improve the algorithm. This feedback mechanism allows the system to refine its judgment criteria dynamically, achieving both high detection accuracy and low false positive rates without requiring manual inspection of every board

Inventive Principle:

Principle #23Feedback

2Ease of operation

If judgment criteria are relaxed to reduce false positives, then ease of operation is improved, but defect detection capability deteriorates

Engineering Contradiction:

Improveautomatic inspection reliabilityVSAvoiddefect detection capability

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The AI system performs self-service by automatically learning from training boards and improving its own detection capabilities without human intervention. This self-service mechanism allows the system to maintain relaxed operation (automatic inspection) while achieving high detection precision through continuous self-improvement based on training data

Inventive Principle:

Principle #25Self-service

3Reliability

If manual inspection stage is added to review false positives, then detection reliability is improved, but inspection time increases

Engineering Contradiction:

Improvedefect detection reliabilityVSAvoidinspection time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system replaces the mechanical manual inspection process with an AI-based automated system. By substituting human inspectors with an AI algorithm trained on visual inspection training boards, the system maintains high detection reliability while eliminating the time loss associated with manual review of false positives

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12626347B2Visual inspection training board for artificial intelligence deep learning

Publication Date: 2026.05.12 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A computing system includes an automatic optical inspection (AOI) and a database that stores images of a plurality of visual inspection (VI) training boards. Each of the visual inspection (VI) training boards includes at least one intended defect. The AOI system includes an artificial intelligence (AI) algorithm implemented in an automatic optical inspection (AOI) system. The AOI system trains the AI algorithm to learn at least one target defect based on the at least one intended defect including in the plurality of VI training boards, analyzes a production printed circuit board (PCB), and determines either a non-defective PCB in response to the AI algorithm determining the production PCB excludes the at least one target defect, or a defective PCB in response to the AI algorithm determining the production PCB includes at least one target defect.