3D-Printing and Machine Learning Control of Soft Ionic Polymer-Metal Composite Actuators

James D Carrico,Tucker Hermans,Kam K Leang,Kwang J Kim

doi:10.1038/s41598-019-53570-y

Abstract

This paper presents a new manufacturing and control paradigm for developing soft ionic polymer-metal composite (IPMC) actuators for soft robotics applications. First, an additive manufacturing method that exploits the fused-filament (3D printing) process is described to overcome challenges with existing methods of creating custom-shaped IPMC actuators. By working with ionomeric precursor material, the 3D-printing process enables the creation of 3D monolithic IPMC devices where ultimately integrated sensors and actuators can be achieved. Second, Bayesian optimization is used as a learning-based control approach to help mitigate complex time-varying dynamic effects in 3D-printed actuators. This approach overcomes the challenges with existing methods where complex models or continuous sensor feedback are needed. The manufacturing and control paradigm is applied to create and control the behavior of example actuators, and subsequently the actuator components are combined to create an example modular reconfigurable IPMC soft crawling robot to demonstrate feasibility. Two hypotheses related to the effectiveness of the machine-learning process are tested. Results show enhancement of actuator performance through machine learning, and the proof-of-concepts can be leveraged for continued advancement of more complex IPMC devices. Emerging challenges are also highlighted.

Highlights

Ionic polymer metal composites (IPMC)s are electroactive-polymer soft actuators with application in biomedical devices and soft robotics
The main contribution of this paper is a paradigm whereby custom-shaped ionic polymer-metal composite (IPMC) actuators can be manufactured as monolithic devices through 3D printing and effective motion control can be achieved through machine learning that avoids the need for modeling the complex behaviors
The validated dynamics model for the leg-body-body-leg configuration was used in simulation to compare the performance of Bayesian optimization to a finite difference policy gradient method to motivate the use of Bayesian optimization as a gait learning method

Summary

Introduction

Ionic polymer metal composites (IPMC)s are electroactive-polymer soft actuators with application in biomedical devices and soft robotics. Fabricating IPMCs typically involves shaping and plating commercially-available sheets and tubular structures of Nafion, Flemion, or Aquivion material This is a laborious and unreliable means of fabricating IPMC-based actuators and devices, and significantly hinders the use of IPMCs in practical applications. These approaches are restrictive in that new designs require new molds, which are time consuming and costly to create In comparison to these methods, free-form-based techniques where structures are created using a layer-by-layer manufacturing process has been proposed[28,29], where layers of Nafion are dispensed into silicone casts and the solvent in the dispersion is allowed to evaporate. Sophisticated, custom-shaped, multi-input-multi-output (MIMO), IPMC-based systems (as may be fabricated through 3D printing) will exhibit coupled nonlinear behavior that may be difficult to model and control To address these challenges, prior works have developed sophisticated control-oriented dynamics models and advanced feedback-control methods to compensate for time-varying and complex dynamic behavior. These features, together with the ability of Bayesian optimization to incorporate prior knowledge to speed up convergence, make it an attractive candidate for use on more complex 3D-printed IPMC actuators and devices

Methods

Results

Conclusion