Mini-LED Mass Transfer Error Compensation Using Neural Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mass transfer methods for Mini-LED chips struggle to accurately position each chip due to complex error sources and high transfer frequency, leading to defects like poor electrical contact and high missing die rates, while current compensation methods only minimize average errors and cannot control individual chip errors.
Innovation Solution
A neural network-based error compensation method that trains a multi-layer neural network with transfer data to predict and pre-compensate for individual Mini-LED chip errors by generating a transfer path, using automated optical re-inspection and data normalization to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If camera-assisted positioning is performed for each Mini-LED chip prior to ejection, then transfer accuracy is improved, but transfer efficiency deteriorates due to high transfer frequency
Solution Approach 1:
The patent performs error compensation in advance by training a neural network model with historical transfer data to establish error characteristics. The model predicts and compensates for systematic errors before actual transfer occurs, eliminating the need for real-time camera positioning of each chip while maintaining accuracy.
Solution Approach 2:
The patent replaces the mechanical camera-assisted positioning system with a neural network-based predictive model. Instead of using optical measurement and physical adjustment for each chip, the system uses data-driven error prediction to pre-correct transfer paths, substituting complex real-time measurement with efficient computational prediction.
2Manufacturing precision
If mathematical models are used to quantify transfer errors, then error control is improved, but model accuracy deteriorates due to complex error sources
Solution Approach 1:
The patent replaces traditional mathematical modeling with a neural network-based data-driven approach. The neural network automatically learns complex error patterns from historical transfer data without requiring explicit mathematical formulations, capturing non-linear relationships and interactions that traditional models miss.
Solution Approach 2:
The patent implements feedback by using actual transfer results to continuously train and update the neural network model. The model learns from real-world error data and improves its prediction accuracy over time, creating a closed-loop system that adapts to changing error characteristics.
3Manufacturing precision
If compensation methods minimize average transfer error, then overall accuracy is improved, but individual chip error control deteriorates
Solution Approach 1:
The patent applies local quality by predicting and compensating for errors of each individual chip rather than applying a uniform compensation to all chips. The neural network model processes each chip's specific characteristics and predicts its unique error, enabling customized compensation for each chip's landing position.
Data Source
AI summary
A neural network-based error compensation method for mass transfer of mini-light-emitting diode (Mini-LED) chips includes the following steps. (S1) An automated optical re-inspection result is obtained as a first result. (S2) The first result is sorted and normalized to obtain a second result. (S3) Nearest-neighbor interpolation is performed on a Mini-LED chip with a transfer status identifier being abnormal, and a differential of path variables of each Mini-LED chip is calculated. (S4) A multi-layer neural network model is defined. A loss function is constructed. A weight of the multi-layer neural network model is updated with the loss function, until no overfitting is observed. (S5) A chip transfer path is generated and input into the multi-layer neural network model to obtain a predicted transfer error value, and mass transfer of the Mini-LED chip is performed based on the predicted transfer error value.