Probe Head Imaging for Automated Defect Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional probe head inspection methods in semiconductor manufacturing are time-consuming, subjective, and prone to human error, leading to inefficient maintenance schedules that either waste resources or risk production errors due to probe degradation.
Innovation Solution
An automated system using computer vision and image processing to monitor probe head condition in real-time, detecting defects such as silicon carbide spikes and mercury residue, generating defect scores, and providing proactive maintenance recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual inspection methods are used to monitor probe head condition, then operators can detect probe degradation, but the inspection process is time-consuming and requires shutdown of equipment
Solution Approach 1:
The patent replaces manual mechanical inspection with an automated optical imaging system. A camera captures images of the probe head, and image processing algorithms automatically analyze the images to detect defects such as silicon carbide spikes and mercury residue. This substitution eliminates the need for manual disassembly and visual inspection, enabling continuous monitoring without equipment shutdown.
Solution Approach 2:
The system creates optical copies (images) of the probe head surface and analyzes these copies to detect degradation. By working with image data rather than the physical probe head itself, the system enables non-contact, non-intrusive inspection that does not require stopping production.
2Manufacturing precision
If frequent manual inspection is performed to ensure measurement accuracy, then probe degradation can be detected early, but production time is lost and resources are wasted
Solution Approach 1:
The automated imaging system enables continuous monitoring of probe head condition during normal production operations. Images are captured and analyzed without interrupting the measurement process, allowing the system to maintain both high measurement accuracy and continuous production throughput. The system can monitor probe degradation trends over time and trigger alerts only when thresholds are exceeded.
Solution Approach 2:
The system performs self-monitoring by automatically capturing images, processing them through algorithms, and generating defect scores without human intervention. This automation eliminates the need for operators to manually inspect probe heads, freeing them to focus on higher-value tasks while maintaining measurement accuracy through continuous automated surveillance.
3Device complexity
If manual visual assessment is used to evaluate probe head condition, then inspection can be performed with simple equipment, but the process is subjective and prone to human error
Solution Approach 1:
The patent replaces subjective human visual assessment with automated image processing algorithms. These algorithms objectively analyze image features such as silicon carbide spike detection, mercury residue identification, and probe tip morphology. The automated analysis eliminates human subjectivity and fatigue-related errors while maintaining equipment simplicity through the use of standard imaging components.
Solution Approach 2:
The system provides objective feedback through quantitative defect scores generated from image analysis. Rather than relying on operator judgment, the system calculates numerical metrics that objectively indicate probe head condition. This feedback mechanism enables consistent, repeatable measurements across different inspection instances and operators.
Data Source
AI summary
The technology disclosed relates to a system and methods for monitoring a probe measurement system, such as monitoring the condition of a probe head or detecting a defect of a probe head. A captured image of a probe head can be processed using an image processing model to extract a feature of the probe head, and the extracted feature can be analyzed to determine a condition indicator and generate a defect score. A defect score can be evaluated based on a pre-defined threshold, and a particular defect score satisfying the pre-defined threshold can indicate that there is a nonconformance associated with the determined indicator, such as a probe head defect. Furthermore, a nonconformance record can be logged for the probe head in response to a defect score satisfying a pre-defined threshold.