HiDeNN-AI is a spin-off from CAMSIM, co-founded by Prof. Liu and Prof. Dong Qian from UT Dallas. HiDeNN-AI is based on three provisional patents technology consisting of an AI platform for scientific and material systems innovation.
The Innovation
HiDeNN-AI, an innovative Mechanistic Data Science (MDS) simulation tool, is sought based on our recently developed technology titled “Hierarchical Deep Learning Neural Networks-Artificial Intelligence (HiDeNN-AI)1-4 platform. Mechanistic Data Science (MDS), a structured methodology for coupling data science methods and tools with mathematical scientific principles (i.e., “mechanistic” principles), is a new concept to solve engineering problems. Our focus will be on the knowledge creation processes and material systems and simulation technology innovation which will provide critical driving forces behind the advancement of Science, Technology, Engineering, Mathematics and the United States workforce. HiDeNN-AI will enable novel simulation technologies that are currently not commercially available.
The Value Proposition
HiDeNN-AI works by combining scientific and mathematical knowledge gained from the creative curation of available data with existing mathematical and scientific principles in combination with deep neural networks (DNN). Through the data collection and generation process, important scientific features are extracted that facilitate highly accurate reduced-order models (ROMs) building, thus significantly reducing the problem dimension. As a result, it enables noteworthy acceleration of high-fidelity computational frameworks to solve previously unsolvable or difficult to solve science and engineering (S&E) problems. In addition, the HiDeNN-AI platform offers superior computational efficiency to existing solution methodologies by utilizing the unique data augmented S&E analysis. As such, HiDeNN-AI opens new avenues to address S&E problems that were previously impossible or highly challenging to solve, thus providing exciting new opportunities in high-performance material/process design and development.
Applications
The expected customers that will benefit greatly from the innovation include:
1) Composite manufacturers and their end users: Polymer matrix composite manufacturers and state of the art innovative material manufactures, including those in the sectors of engineered materials, transportation, medical devices, and consumer electronics, such as 3M, Corning, Dow Chemical, BASF, Apple, Caterpillar, Goodyear, Ford Motors, General Motors, General Electric, Siemens, Firestone, Medtronic, and competitors, pose to largely benefit from the HiDeNN-AI platform.
2) Advanced manufacturing: With the metal additive manufacturing sector expected to grow into a $10 B industry within the next decade, HiDeNN-AI will be a lucrative tool to maintain global competitive advantage.
3) Software companies focusing on product design: Companies such as ANSYS, Dassault systems, Hexagon, Autodesk, and Altair show immense interest in incorporating data-driven tools into their current simulation platforms and are looking for partnerships.
Given the significant demands called for by the industries, mechanistic data science software with physics-based machine learning reduced order models is a powerful and needed tool for the unmet applications in lightweight, high-performance, and multifunctional materials design and manufacturing.
BRIEF INTRODUCTION
Please feel free to download our brief Slide Deck to introduce you to what the HiDeNN-AI platform can do.