How Machine Learning Will Revolutionize Electrochemical Sciences

Abstract

Electrochemical systems function via interconversion of electric charge and chemical species and represent promising technologies for our cleaner, more sustainable future. However, their development time is fundamentally limited by our ability to identify new materials and understand their electrochemical response. To shorten this time frame, we need to switch from the trial-and-error approach of finding useful materials to a more selective process by leveraging model predictions. Machine learning (ML) offers data-driven predictions and can be helpful. Herein we ask if ML can revolutionize the development cycle from decades to a few years. We outline the necessary characteristics of such ML implementations. Instead of enumerating various ML algorithms, we discuss scientific questions about the electrochemical systems to which ML can contribute.

A Perspective on Li/S Battery Design: Modeling and Development Approaches

“…ML may be the key to unlocking the full potential of MSMs. They could be used to couple multiscale simulations, accelerate model development, and ensure experimental validity [131]. Recently, Kolodziejczyk et al demonstrated the effectiveness of ML in MSM development.…”

Supervised Learning of Synthetic Big Data for Li‐Ion Battery Degradation Diagnosis

“…Recently, there has been interest in the community in combining the physical grounding and interpretability of mechanistic models with the adaptability and ease of evaluation of empirical ML models [21,22] . Specifically, supplementing or replacing experimental training data for ML models with synthetic training data generated by a physics based or mechanistic model has been proposed as a solution to the data limitation problem [23] …”

Machine Learning 3D-Resolved Prediction of Electrolyte Infiltration in Battery Porous Electrodes

“…Meanwhile, Artificial Intelligence (AI) has seen a tremendous rise in the last decade, becoming essential for modern industry and finance, among many other fields. 18 In LIBs, machine learning (ML) techniques have enabled tools that significantly reduce the slow time frames related to trial-and-error approaches or physics-based simulations for faster and more efficient data assessment. [19][20][21][22] Deep Neural Networks (DNNs) are the most popular technique in the AI field due to the good performances they show for modeling complex data structures with many non-linear relationships.…”

“…18 In LIBs, machine learning (ML) techniques have enabled tools that significantly reduce the slow time frames related to trial-and-error approaches or physics-based simulations for faster and more efficient data assessment. [19][20][21][22] Deep Neural Networks (DNNs) are the most popular technique in the AI field due to the good performances they show for modeling complex data structures with many non-linear relationships. [23][24] Particularly, Convolutional Neural Networks (CNNs), a type of DNN, are a perfect example, having outstanding performances in different applications involving many types of data such as images-to-images translations, image classification, or autonomous driving.…”