Artificial Intelligence in Computational Mechanics and Biomechanics

This work aims to deliver a brief presentation and evolution of Artificial intelligence (AI) and its potentially most suited methodologies for Computational Mechanics and Biomechanics (CM&B) applications, such as Machine Learning (ML), Pattern Recognition (PR), Deep Learning (DL) and DL using artificial Neural Networks (NN). Afterwards, since DL using artificial NN is the AI methodology mostly used in CM&B, this methodology applied to CM&B problems is described with more detail. In order to show the evolution of AI methodologies in CM&B, a large document search was performed in the academic database Web of Science. Thus, peer-reviewed relevant articles addressing the topics of ML, PR, DL and NN combined with CMB&B were selected for a quantitative analysis. The results confirm that DL using artificial NN is in fact the most used AI methodology in both CM&B. Furthermore, it was found that research using DL using artificial NN combined with the finite element method is growing much faster than any other methodology combination. This work shows the inevitable growth of AI, which will accelerate the computation of today’s demanding problems and will allow the simulation of highly complex problems beyond the competence of existing rigid computational methodologies. AI offers the opportunity to expand the traditional application fields of CM&B, which will change its paradigm in a very near future.