Autonomy for DOE Simulations

PI Allan Grosvenor, MSBAI — dba Microsurgeonbot Inc
Project Summary

The successful outcome of this research will achieve new scientific outcomes that demonstrate the latest potential for HPC, simulation, and AI to hyper-enable engineers to solve humanity's largest challenges, and the business implications of our success in this effort are represented by the $30 trillion opportunity described above.

Project Description

GURU, is an autonomous system that drives expert workflows in software, to remove the barriers that prohibit engineers from achieving the promise of simulation and High Performance Computing (HPC). MSBAI built and demonstrated GURU, reducing Virtual Prototype modeling and simulation setup time from hours to minutes for a range of engineering applications. This latest effort focuses on expanding the flexibility and usability of the system, by adding hybrid interaction, so users can type or speak requests, as well as gesturing and even sketching. Advancements to GURU's learning engine (the skills agent factory ), and new R&D will improve the speed and flexibility of generating new high quality, readily transportable, geometrical Computer Aided Design (CAD) files. 

A broad range of industries, from energy, automotive, and aerospace, to telecommunications, medical devices, and consumer products, expect a third of their sales to come from new products, which means that $30 trillion in revenues depends upon successful product development. In response to this critical need, organizations have increasingly adopted simulation as a standard practice, yet only 5% of these industries have fully integrated Artificial Intelligence (AI) into their operations. The setup of simulations remains a laborious process, potentially demanding months or even years for users to fully master. This challenge poses a barrier to achieving the Department of Energy's Exascale Computing Project (ECP) Vision. One of the longest-standing barriers to productivity in engineering utilization of computational physics simulations, as well as manufacturing, is the preparation and exchange of geometrical data. Data, quality, or functionality are routinely lost when converting, importing, or exporting CAD files which often causes engineers to lose enormous amounts of time and prohibits them from solving design and manufacturing challenges. Humanity has an opportunity to enter a new renaissance in creative and engineering design, as we bring the power of validated specialized modeling and simulation software together with HPC to scale up their utilization, driven by the latest AI methodologies to enable the mass adoption of these tools. The successful outcome of this research will achieve new scientific outcomes that demonstrate the latest potential for HPC, simulation, and AI to hyper-enable engineers to solve humanity's largest challenges, and the business implications of our success in this effort are represented by the $30 trillion opportunity described above.

Allocations