An Agentic Orchestration of Atomistic Simulations
For computational materials scientists, this automates complex simulation workflows, reducing human effort and trial-and-error, but the approach is incremental as it applies existing agent technology to a specific domain.
This work introduces an agent-based system within the URSA framework that automates atomistic simulation workflows using LAMMPS, reducing manual intervention and improving reproducibility. The system autonomously selects potentials, runs simulations, and performs error recovery, with reliability benchmarked against the LAVA toolkit.
Atomistic simulations are central to materials design, but their execution involves complex, multi-step workflows that require significant human expertise. Here, we present an agent-based system embedded within the URSA (Universal Research and Scientific Agent) framework that automates the design, execution, and validation of atomistic simulations, demonstrated using the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) tool. Our system autonomously selects interatomic potentials, constructs and runs simulations, and performs iterative error recovery within a closed-loop workflow. We evaluate the scientific reliability of the agent by benchmarking its outputs against LAVA, a high-throughput toolkit for LAMMPS and the Vienna Ab initio Simulation Package (VASP) calculations. Our framework reduces manual intervention and trial-and-error, thereby improving the rigor, reproducibility, and scalability of atomistic modeling.