Related Experiment Video
Updated: Jan 3, 2026
14:42
Fabrication of Carbon-Based Ionic Electromechanically Active Soft Actuators
Published on: April 25, 2020
8.7K
3D-Printing and Machine Learning Control of Soft Ionic Polymer-Metal Composite Actuators.
James D Carrico1, Tucker Hermans2, Kwang J Kim3
1University of Mary, School of Engineering, Bismarck, ND, 58504, USA.
Scientific Reports
|November 27, 2019
Summary
This study introduces 3D printing for soft ionic polymer-metal composite (IPMC) actuators and uses Bayesian optimization for control. This enhances actuator performance and enables novel soft robotic designs.
Area of Science:
- Robotics
- Materials Science
- Additive Manufacturing
Background:
- Traditional soft ionic polymer-metal composite (IPMC) actuator fabrication faces challenges in customization and integration.
- Existing control methods for IPMC actuators often require complex models or continuous sensor feedback, limiting their adaptability.
Purpose of the Study:
- To present a novel manufacturing and control paradigm for soft IPMC actuators.
- To enable the creation of custom-shaped, integrated soft actuators using additive manufacturing.
- To improve the control of these actuators by mitigating complex dynamic effects using machine learning.
Main Methods:
- Additive manufacturing, specifically fused-filament (3D printing), was employed using ionomeric precursor material to create monolithic IPMC devices.
- Bayesian optimization, a learning-based control approach, was utilized to manage time-varying dynamic effects in the 3D-printed actuators.
- The developed manufacturing and control paradigm was applied to create and control example actuators, culminating in a modular reconfigurable IPMC soft crawling robot.
Main Results:
- The 3D printing method successfully produced custom-shaped, monolithic IPMC actuators with potential for integrated sensors and actuators.
- Bayesian optimization demonstrated enhanced actuator performance by effectively mitigating complex dynamic effects without requiring intricate models or constant sensor feedback.
- The feasibility of the paradigm was proven through the construction and control of a soft crawling robot composed of the developed IPMC actuators.
Conclusions:
- The presented manufacturing and control paradigm offers a viable approach for advancing soft IPMC actuator development.
- The integration of 3D printing and machine learning-based control significantly enhances actuator performance and adaptability.
- The proof-of-concept provides a foundation for developing more complex IPMC-based soft robotic systems.
