ABSTRACT

In this work, a novel analysis approach using an artificial intelligence (AI) framework that automatically helps identify the drivers of a formulation from lab measurements was showcased. This AI framework builds and optimizes models in the form of physics-based equations from small amounts of measurement data. With this approach, the user can overcome the large data requirements of machine learning while building tailored models that outperform traditional analytical and statistical tools.

The effectiveness of this modeling framework in helping scientists reduce uncertainty early in the experiment process was demonstrated. This solution allows a significant reduction in the number of experiments required to achieve an optimal formulation. This was accomplished by generating new, custom models from existing data and well-known equations in electrochemistry. Then, these models were used to predict or hypothesize the performance of unseen formulations by altering their control parameters. This study showed the accuracy of these predictions by calculating its error against unseen measurement data.

INTRODUCTION

In most engineering and scientific applications, machine learning (ML) or artificial intelligence (AI) methods in general, are primarily oriented to design a statistical/heuristic procedure to predict the outcome of a system under new conditions. This mechanism aims at exploring non-evident correlations between inputs and outputs that are embedded in the data. However, a large body of this effort relies on black-box function approximations (e.g., neural networks) that have shown limitations to elucidate additional insights from the underlying physical process that generated the data. Thus, this type of knowledge is generated in a data-driven manner without fully explaining the physics governing the problem.

In contrast, a mechanistic modeling approach proceeds from the starting point of well-established scientific laws and axioms, produced by a means of logical deduction physics-based concepts explaining a phenomenon and measurements of it. Typically, the mechanistic modeling approach describes causal mechanisms by simplified mathematical formulations, while the ML/AI approach seeks to establish a statistical relationship between inputs and outputs.

Both approaches are common and, in some cases, offer competitive views across a wide range of electrochemical applications beyond corrosion (e.g., biomedical devices, semiconductors, sensors, batteries, fuel cells, coatings, and imaging). In the present work, machine learning and mechanistic modeling are complementing each other to automate the development and evaluation of corrosion inhibitor actives by a means of electrochemical procedures.

This content is only available via PDF.
You can access this article if you purchase or spend a download.