Exploration of deep operator networks for predicting the piezoionic effect

Abstract

The piezoionic effect holds significant promise for revolutionizing biomedical electronics and ionic skins. However, modeling this multiphysics phenomenon remains challenging due to its high complexity and computational limitations. To address this problem, this study pioneers the application of deep operator networks to effectively model the time-dependent piezoionic effect. By leveraging a data-driven approach, our model significantly reduces computational time compared to traditional finite element analysis (FEA). In particular, we trained a DeepONet using a comprehensive dataset generated through FEA calibrated to experimental data. Through rigorous testing with step responses, slow-changing forces, and dynamic-changing forces, we show that the model captures the intricate temporal dynamics of the piezoionic effect in both the horizontal and vertical planes. This capability offers a powerful tool for real-time analysis of piezoionic phenomena, contributing to simplifying the design of tactile interfaces and potentially complementing existing tactile imaging technologies.