# The Joint Automated Repository for Various Integrated Simulations (JARVIS) for data-driven materials design

**Authors:** Kamal Choudhary, Kevin F. Garrity, Andrew C. E. Reid, Brian DeCost, Adam J. Biacchi, Angela R. Hight Walker, Zachary Trautt, Jason Hattrick-Simpers, A. Gilad Kusne, Andrea Centrone, Albert Davydov, Jie Jiang, Ruth Pachter, Gowoon Cheon, Evan Reed, Ankit Agrawal, Xiaofeng Qian, Vinit Sharma, Houlong Zhuang, Sergei V. Kalinin  
**arXiv:** [2007.01831](https://arxiv.org/abs/2007.01831)  
**Date:** 2020-07-03  
**Summary page:** https://www.summarizepaper.com/en/paper/2007.01831v2/

## Key points

- The Joint Automated Repository for Various Integrated Simulations (JARVIS) is an infrastructure designed to accelerate materials discovery and design.
- JARVIS uses density functional theory (DFT), classical force-fields (FF), and machine learning (ML) techniques.
- JARVIS aims to reduce the cost and time associated with materials discovery, optimization, and deployment.
- Major components of JARVIS include JARVIS-DFT, JARVIS-FF, JARVIS-ML, and JARVIS-Tools.
- Currently contains 40,000 materials and 1 million calculated properties in JARVIS-DFT; 1,500 materials and 110 force-fields in JARVIS-FF; as well as 25 ML models for material-property predictions in JARVIS-ML.
- Provides scripts and workflows for running simulations and analyzing results through its component -JARVISTools-
- Incorporates a rigorous evaluation process by comparing computational data with experimental or high-fidelity computational methods whenever possible.
- Supports various other technologically important applications within the data-driven materials design paradigm.
- Team behind the project consists of researchers from multiple institutions including NIST, Theiss Research, University of Maryland Air Force Research Laboratory Stanford University Northwestern University Texas A&M University Arizona State University Oak Ridge National Laboratory Los Alamos National Lab Virginia Tech Rutgers University.
- Serves as a valuable resource for researchers in the field of materials science by providing access to extensive datasets and tools.

## Summary

The Joint Automated Repository for Various Integrated Simulations (JARVIS) is an integrated infrastructure designed to accelerate materials discovery and design through the use of density functional theory (DFT), classical force-fields (FF), and machine learning (ML) techniques. Built upon the principles of the Materials Genome Initiative (MGI), JARVIS aims to reduce the cost and time associated with materials discovery, optimization, and deployment by providing open-access databases and tools. The major components of JARVIS include JARVIS-DFT, JARVIS-FF, JARVIS-ML, and JARVIS-Tools. Currently, it contains 40,000 materials and 1 million calculated properties in JARVIS-DFT; 1,500 materials and 110 force-fields in JARVIS-FF; as well as 25 ML models for material-property predictions in JARVIS-ML. These datasets are continuously expanding to encompass more diverse materials. JARVIS also provides scripts and workflows for running simulations and analyzing results through its component -JARVISTools-. Additionally, a rigorous evaluation process is incorporated into the infrastructure by comparing computational data with experimental or high-fidelity computational methods whenever possible to assess prediction errors or uncertainties. Its flexible nature allows it to support various other technologically important applications within the data-driven materials design paradigm. The team behind this project consists of researchers from multiple institutions including NIST, Theiss Research, University of Maryland Air Force Research Laboratory Stanford University Northwestern University Texas A&M University Arizona State University Oak Ridge National Laboratory Los Alamos National Lab Virginia Tech Rutgers University. In conclusion, JARVIS serves as a valuable resource for researchers in the field of materials science by providing access to extensive datasets and tools that facilitate efficient materials discovery and design processes which are publicly available on its website for utilization in their own investigations.