Fabric Printing Liquid Selection Using Pre-Print Observation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing printing systems struggle to maintain consistent image quality due to variations in fabric permeability, leading to blurring and uneven distribution of printing liquids, as they do not adequately account for the specific properties of different fabrics.
Innovation Solution
A printing system comprising a first system with a printing unit and observation unit, and a second system connected via a network, which utilizes machine learning to recommend a suitable printing liquid based on pre-print observation data using a trained model to optimize image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If printing liquid is ejected onto fabric without considering fabric permeability, then printing process is simple, but image quality becomes inconsistent due to blurring and uneven distribution
Solution Approach 1:
The system performs preliminary observation of fabric permeability using the observation unit before printing. The acquiring unit obtains fabric information in advance, and the model predicts suitable printing liquid types beforehand, allowing the printing process to proceed with pre-determined optimal parameters, thus ensuring image quality without adding complexity to the actual printing operation.
Solution Approach 2:
A machine learning model serves as an intermediary between the fabric observation data and the printing liquid selection. The model takes fabric permeability characteristics as input and outputs recommended printing liquid types, acting as a smart mediator that translates fabric properties into optimal printing parameters without requiring direct complex interaction between the printing unit and fabric analysis.
2Manufacturing precision
If printing liquid type is selected manually to match fabric permeability, then image quality can be maintained, but operator skill and time consumption increase
Solution Approach 1:
The system implements self-service by automatically observing fabric properties through the observation unit, acquiring fabric information through the acquiring unit, and autonomously determining the optimal printing liquid type using the machine learning model. This eliminates the need for operator intervention in liquid selection, saving time while maintaining consistent image quality based on actual fabric characteristics.
Solution Approach 2:
The system establishes a feedback loop where the observation unit continuously monitors fabric permeability, the acquiring unit collects this data, and the model adjusts printing liquid recommendations based on real-time fabric information. This closed-loop feedback mechanism ensures that the printing liquid selection is always optimized for the current fabric being printed, maintaining high image quality without manual intervention.
3Ease of operation
If standard printing liquid is used for all fabric types, then system operation is simple, but permeability variations cause blurring and poor liquid setting
Solution Approach 1:
The system transitions from a static, one-size-fits-all printing liquid selection to a dynamic, adaptive approach. The observation unit dynamically assesses fabric permeability for each printing task, and the machine learning model dynamically recommends the appropriate printing liquid type based on real-time fabric characteristics. This dynamic adaptation maintains operational simplicity while significantly improving liquid setting performance and reliability across different fabric types.
Data Source
Figure 1
Figure 2
AI summary
A printing system includes at least one first system including a printing unit configured to eject a printing liquid onto a fabric to perform printing on the fabric, and an observation unit configured to observe the fabric to acquire observation data, and a second system configured to communicate with the at least one first system through a network, the second system including an acquiring unit configured to acquire pre-print observation data acquired by the observation unit observing the fabric before printing, a storage unit configured to store a model trained through machine learning, the model being configured to output a physical property parameter of a recommended liquid that is printing liquid recommended for the fabric represented by the pre-print observation data when the pre-print observation data is input, and an output unit configured to output, to the at least one first system, recommended data representing the physical property parameter output by the model.