Automated FIB Sample Preparation Using Image Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing complexity and miniaturization of devices make manual production of TEM analysis samples using charged particle beams labor-intensive and prone to errors, limiting the ability of unskilled technicians to accurately determine the working end position, thus necessitating a more efficient and automated process.
Innovation Solution
An automatic working apparatus that includes a structure information obtaining unit, a working end position obtaining unit, an image obtaining unit, and a determining unit to compare pre-working and real-time images of the sample to determine if the charged particle beam has reached the correct end position, allowing for automated sample production without human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual working is performed by an operator to leave only one desired structural body, then the working can be performed with flexibility, but the work takes a very long time and increases the risk of errors
Solution Approach 1:
The system performs automatic working end position determination by itself using image recognition and comparison algorithms, eliminating the need for operator intervention. The apparatus autonomously compares obtained images with pre-stored structure information to determine when working should end, thereby increasing productivity while maintaining accuracy.
Solution Approach 2:
The patent replaces manual visual inspection and decision-making by operators with an automated image processing system. The determining unit uses computational algorithms to compare images and identify structural bodies, substituting human cognitive processes with mechanical/optical systems that operate faster and without fatigue.
2Ease of manufacture
If an unskilled technician performs manual working, then labor availability increases, but the work takes a very long time and accuracy decreases
Solution Approach 1:
The system continuously obtains images during working, compares them with pre-stored structure information, and uses the comparison results to determine the working end position. This closed-loop feedback mechanism ensures that even unskilled technicians can achieve high precision by relying on the automated determination system rather than manual skill.
Solution Approach 2:
The patent stores pre-acquired images representing the complete structure information as a reference template. During working, the system compares real-time images against this stored copy to identify when the desired structural body remains, eliminating the need for technicians to have expert knowledge of the target structure.
3Manufacturing precision
If manual working is performed with careful observation, then accuracy can be maintained, but the work takes a very long time
Solution Approach 1:
The image obtaining unit continuously captures images during the working process without interruption, and the determining unit continuously compares these images with stored structure information. This continuous automated monitoring enables rapid determination of the working end position, maintaining high accuracy while dramatically reducing the time required compared to intermittent manual observation.
Solution Approach 2:
The patent replaces slow manual visual observation and decision-making with automated image processing systems that can analyze multiple images per second. The computational comparison of images with stored structure information occurs much faster than human visual processing, thereby reducing time loss while maintaining or improving accuracy.
4Productivity
If automated working is implemented, then productivity increases, but the ability to handle complex structures may be reduced
Solution Approach 1:
The system enhances its adaptability to complex structures by changing the parameters of the image processing and pattern recognition algorithms. By adjusting sensitivity thresholds, comparison criteria, and structural recognition parameters, the automated system can accommodate varying degrees of structural complexity while maintaining high productivity through rapid automated determination.
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
The apparatus enables accurate and efficient determination of the working end position for charged particle beams, reducing labor and skill dependency, thereby improving the production efficiency and reliability of TEM analysis samples.
Implementation Method 1
irradiating the sample with a charged particle beam
Implementation Method 2
irradiating the sample with a charged particle beam
Implementation Method 3
obtain a working surface image, the working surface image being obtained by imaging a working surface that appears at a position where the sample is irradiated with the charged particle beam
Data Source
AI summary
This automatic processing device for fabricating a sample piece from a sample by irradiating the sample with a charged particle beam is provided with: a structural information acquiring unit which acquires structural information indicating the structure of the sample before processing; a processing termination position acquiring unit which acquires termination position specifying information specifying a processing termination position corresponding to the structure of the sample; an image acquiring unit which acquires a processed surface image in which a processed surface appearing at the position at which the sample has been irradiated by the charged particle beam is captured; and a determining unit which determines whether the position of the processing by the charged particle beam has reached the termination position, on the basis of a comparison between the structural information acquired by the structural information acquiring unit and the processed surface image acquired by the image acquiring unit.