Autonomous Inkjet Printing Optimization via Active Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current inkjet printing systems require time-consuming and resource-intensive calibration processes to ensure consistent and reliable jetting of ink materials, particularly due to challenges in predicting the acoustics within print heads and the need for extensive experimental data collection, which is costly and inefficient, especially when using expensive materials like gold or biomaterials.
Innovation Solution
The implementation of active machine learning with model selection to autonomously control printing systems, reducing the number of experiments required for generating jettability diagrams by iteratively determining decision boundaries and selecting informative data points, allowing for fully automated loop operations without human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional trial-and-error calibration methods are used to ensure reliable jetting, then jetting reliability is improved, but time consumption and material cost increase significantly
Solution Approach 1:
The system performs self-calibration by automatically determining decision boundaries and selecting informative data points through machine learning algorithms, eliminating the need for manual trial-and-error calibration by operators
Solution Approach 2:
The system uses feedback from jetting experiments to iteratively improve the classification model, using the results of each experiment to refine the decision boundary and guide subsequent experiments, thereby reducing total calibration time while maintaining reliability
2Manufacturing precision
If extensive experimental data collection is performed to map jetting zones, then manufacturing precision is improved, but material consumption and cost increase
Solution Approach 1:
The system performs only the necessary experiments to achieve sufficient precision by using active learning to select the most informative data points, avoiding excessive experimentation and reducing material consumption while maintaining adequate jetting zone definition
Solution Approach 2:
The system changes the approach from exhaustive parameter sampling to intelligent parameter selection based on decision boundary identification, using machine learning to determine which parameter combinations are most informative for defining jetting zones
3Adaptability or versatility
If manual calibration processes are used for new ink-print head combinations, then adaptability is improved, but productivity decreases due to time-consuming operations
Solution Approach 1:
The system replaces manual mechanical calibration operations with automated machine learning-based classification, using algorithms to assess compatibility between inks and print heads, thereby maintaining adaptability while significantly increasing productivity
Solution Approach 2:
The system introduces a machine learning classification model as an intermediary between ink-print head combinations and jetting performance assessment, automatically determining compatibility without requiring manual trial-and-error testing
Data Source
AI summary
Active machine learning with model selection is used to control a printing system to efficiently, accurately, and autonomously predict jettability diagrams for different print head and ink combinations. A print system may be controlled using active machine learning to efficiently and accurately predict a jetting behavior for different combinations of ink and inkjet print heads at different settings/values for operating parameters of the print heads, such as firing voltage, pulse width/dwell time, frequency, ramp up, ramp down, meniscus pressure, heating rate, hold time, transducers' positions, geometries, and heating power, and/or other operating parameters of the print heads.