Towards artificial intelligence at scale in the chemical industry

Abstract

In the Industry 4.0 era, the chemical industry is embracing broad adoption of artificial intelligence (AI) and machine learning (ML) methods. This article provides a holistic view of how the industry is transforming digitally towards AI at scale. First, a historical perspective on how the industry used AI to aid humans in better decision‐making is shown. Then state‐of‐the‐art AI research addressing industrial needs on reliability and safety, process optimization, supply chain, material discovery, and reaction engineering is highlighted. Finally, a vision of the plant of the future is illustrated with critical components of AI‐ready culture, model life cycle management, and renewed role of humans in chemical manufacturing.

Distance Matrix Patterns for Visual and Interpretable Process Data Analytics

“…Alternatively, the eighth column could be used, given that the distance matrix is symmetric (D ij = D ji ). It is worth noting that the coordinates (8, 3) and (3,8) appear darker in color, indicating that these samples are the farthest apart from each other. Additionally, the main diagonal of the matrix is colored white, as the distance from a sample to itself is zero.…”

“…The fault detection index is based on the average dissimilarity of a given distance matrix in relation to all others. The dissimilarity metric selected for this work is the Frobenius norm, which has the form (3) The Frobenius norm is a measure of the Euclidean distance between matrices when the matrix space is considered Euclidean. 49 The fault detection index for a given data window is thus defined as the average Frobenius norm of the distance matrix that represents the present window in relation to all other matrices that compose the training database {D 1 ,•••D K }.…”

“…Growing demands for process operational quality, safety, and productivity have been motivating a growing interest in research on process data analytics. This interest grows synergistically with advances in computational capacity and plant instrumentation, contributing to an exponential expansion of the field. − …”

Challenges and Opportunities of Implementing Data Fusion in Process Analytical Technology—A Review

“…The more powerful use of these tools for the materialization of autonomous systems stands in the development of interconnections between sensors and equipment from all unit operations [ 178 ]. The DF techniques combined with AI or machine learning can support the decision-making based on key performance indicators in industrial chemical plants [ 179 ]. Platforms are available, in which in silico development and optimization are performed by data-driven models and digital twins for pharmaceutical systems [ 180 ].…”

A Transformer-Convolutional Neural Network Based Framework for Predicting Ionic Liquid Properties

“…16 In addition to the above methods, machine learning (ML) techniques for molecular property prediction have recently gained in popularity in cheminformatics and promoted broad applications of data-driven models in chemical engineering studies. [17][18][19][20][21] With the availability of IL property databases such as the ILThermo 22 , there has been a sharp rise in the use of data-driven methods for modelling IL properties. [23][24][25][26][27][28][29][30] Among these ML models, different types of molecular descriptors have been used for IL representation.…”