Towards industry-ready additive manufacturing: AI-enabled closed-loop control for 3D melt electrowriting
Pawel Mieszczanek1,2, Peter Corke2, Courosh Mehanian3,4
1ARC Training Centre in Additive Biomanufacturing, Queensland University of Technology, Brisbane, QLD, Australia.
Communications Engineering
|November 5, 2024
Summary
Melt electrowriting (MEW) 3D printing is improved using machine learning and computer vision. This closed-loop control enhances reproducibility and streamlines the complex process for industrial applications.
Area of Science:
- Biomedical Engineering
- Materials Science
- Robotics
Background:
- Melt electrowriting (MEW) is a high-resolution 3D printing technique with applications in regenerative medicine and soft robotics.
- Transitioning MEW from research to industry is hindered by slow experimentation, low throughput, and poor reproducibility due to its complex, nonlinear nature.
Purpose of the Study:
- To overcome the challenges hindering industrial adoption of MEW.
- To develop a closed-loop control system for MEW using artificial intelligence.
Main Methods:
- Real-time monitoring of the MEW jet using computer vision and machine learning.
- Development of an automated data collection methodology to accelerate experimental timelines.
- Implementation of a feedforward neural network with optimization and feedback loops for closed-loop control.
Main Results:
- Significantly reduced experimental time from days to hours through automated data collection.
- Demonstrated successful closed-loop control of the MEW process, ensuring reproducibility of printed parts.
- Streamlined the operation of the nonlinear MEW 3D printing technology.
Conclusions:
- Machine learning and computer vision offer a viable solution for real-time monitoring and control of MEW.
- Closed-loop control enhances the reproducibility and efficiency of MEW, paving the way for industrial applications.
- AI-driven approaches are crucial for advancing complex 3D printing technologies like MEW.
