Related Experiment Video
Updated: Aug 5, 2026
04:32
Quantitative Characterization of Liquid Photosensitive Bioink Properties for Continuous Digital Light Processing Based Printing
Published on: April 14, 2023
Measurement-Informed Deep Learning for Real-Time Monitoring of Hydrodynamic Drying Dynamics in Inkjet-Printed Display
Sung Jun Park1, Donggeun Park2, Jin Hong Park1
1Department of Semiconductor Engineering, Gyeongsang National University, Jinjudae-ro 501 Beon-gil, Jinju-si, Gyeongsangnam-do 52828, Republic of Korea.
ACS Applied Materials & Interfaces
|July 27, 2026
Summary
This study introduces a novel AI framework for real-time monitoring of ZnO nanoparticle ink drying stages in inkjet printing. The system accurately classifies drying phases using low-resolution images, enabling efficient quality control.
Area of Science:
- Materials Science
- Nanotechnology
- Artificial Intelligence
Background:
- Real-time monitoring of thin-film fabrication is crucial for quality control in inkjet printing.
- Understanding drying dynamics of nanoparticle inks influences film properties and device performance.
Purpose of the Study:
- To develop a classification framework for ZnO nanoparticle ink drying stages using low-resolution optical images.
- To establish a computationally efficient method for real-time quality control in inkjet printing.
Main Methods:
- Correlating 3D surface profiles with visual textures to define drying stages.
- Training a lightweight convolutional neural network (CNN) on downsampled grayscale images.
- Utilizing explainable AI (XAI) techniques like Grad-CAM for model interpretation.
Main Results:
- Defined three drying stages for binary solvent inks and two for single-solvent systems.
- Achieved 99.4% test accuracy with a lightweight CNN trained on 115 × 40 pixel images.
- Demonstrated computationally efficient inference (0.64 ms/image) with a 16.4-fold reduction in cost.
Conclusions:
- The developed image-only approach enables reliable and efficient real-time process monitoring for ZnO nanoparticle inkjet printing.
- The lightweight CNN autonomously learns critical visual features for drying stage classification.
- This work provides a practical pathway for real-time quality control in nanoparticle-based additive manufacturing.
