Laser Reflection Monitoring for Real-Time Hole Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional laser processing systems face challenges in detecting defects in real time during the processing of targets, leading to high costs due to post-processing quality inspection and difficulty in correcting or reworking defective molds.
Innovation Solution
A method and apparatus for monitoring laser processing that involves irradiating a laser to a target, receiving reflected light, identifying reflection patterns, and comparing them with previously stored learning patterns to detect defects using an unsupervised autoencoder, thereby enabling real-time defect detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If post-processing quality inspection is performed, then defect detection is achieved, but processing time is increased and costs are elevated
Solution Approach 1:
The system performs preliminary defect detection during the laser processing operation itself by continuously monitoring reflected light patterns, rather than waiting until post-processing inspection. The learning pattern is established beforehand for each target type, enabling real-time comparison and immediate defect identification during processing, thus eliminating the need for separate post-processing inspection time
Solution Approach 2:
The defect detection process runs continuously throughout the laser processing operation without interruption. The system continuously receives reflected light, compares it against the learning pattern, and identifies defects in real-time, making the inspection process concurrent with manufacturing rather than sequential, thereby eliminating additional inspection time
2Productivity
If real-time defect detection is implemented, then separate inspection processes are eliminated, but device complexity increases
Solution Approach 1:
The monitoring system is designed to handle multiple target types by storing and comparing against different learning patterns for each target type. The same hardware infrastructure (light receiver, processor, memory) serves all target types, making the system multi-functional rather than requiring separate inspection systems for each target type, thus limiting the increase in device complexity
Solution Approach 2:
The system creates a digital copy (learning pattern) of the expected reflected light pattern for each target type and stores it in memory. During processing, the actual reflected light pattern is compared against this digital copy to identify defects. This copying approach enables real-time detection using simple comparison operations rather than complex analysis, maintaining system simplicity while achieving high productivity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables real-time defect detection during processing, reducing the need for separate inspection processes and minimizing defect detection time and costs.
Implementation Method 1
receiving reflected light that is the laser reflected by the target
Data Source
AI summary
The present invention relates to a laser processing monitoring method and apparatus, the method comprising the steps of: irradiating a laser to a target according to a start signal for target processing, receiving reflected light that is the laser reflected by the target, identifying a reflection amount pattern for the reflected light, and detecting whether a hole generated by the laser irradiation is defective based on the reflection amount pattern. In addition, the present invention may have other embodiments.