arXiv subjects

Ryley G. Hill

Publications and source records attributed to Ryley G. Hill.

3 recordsLinked to original sources

arXiv2026

TREMORS: An Agentic Assistant for Multi-Datacenter Seismic Data Acquisition

Seismology increasingly depends upon large data retrieval across multi-datacenter platforms. Yet, data procurement often demands domain expertise, user burden, and is difficult to reproduce. As archives continue to grow, translating scientific intent into structured workflows that operate across multiple repositories and produce high quality, AI ready data is becoming an increasingly urgent challenge. We present TREMORS (Text Referenced Event Mapping and Output Renderer for Seismographs), an agentic framework that uses large language model reasoning within a constrained execution graph to automate seismic data retrieval. TREMORS translates natural language queries into a structured intermediate schema, which drives execution through a constrained LangGraph workflow. Example workflows demonstrate support for both event-based and continuous waveform acquisition. The framework is designed to extend across heterogeneous multi-datacenter systems through a portable schema and modular workflow components. This work positions agentic workflows as the catalyst that will connect scientific intent with distributed seismic data systems, enabling a future of reproducible, data-driven inquiry while reducing the burden of routine acquisition tasks.

physics.geo-ph↗

arXiv2026

An Agentic Orchestration of Atomistic Simulations

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.

cs.AI↗

arXiv2025

A Foundation Model for Material Fracture Prediction

Accurately predicting when and how materials fail is critical to designing safe, reliable structures, mechanical systems, and engineered components that operate under stress. Yet, fracture behavior remains difficult to model across the diversity of materials, geometries, and loading conditions in real-world applications. While machine learning (ML) methods show promise, most models are trained on narrow datasets, lack robustness, and struggle to generalize. Meanwhile, physics-based simulators offer high-fidelity predictions but are fragmented across specialized methods and require substantial high-performance computing resources to explore the input space. To address these limitations, we present a data-driven foundation model for fracture prediction, a transformer-based architecture that operates across simulators, a wide range of materials (including plastic-bonded explosives, steel, aluminum, shale, and tungsten), and diverse loading conditions. The model supports both structured and unstructured meshes, combining them with large language model embeddings of textual input decks specifying material properties, boundary conditions, and solver settings. This multimodal input design enables flexible adaptation across simulation scenarios without changes to the model architecture. The trained model can be fine-tuned with minimal data on diverse downstream tasks, including time-to-failure estimation, modeling fracture evolution, and adapting to combined finite-discrete element method simulations. It also generalizes to unseen materials such as titanium and concrete, requiring as few as a single sample, dramatically reducing data needs compared to standard ML. Our results show that fracture prediction can be unified under a single model architecture, offering a scalable, extensible alternative to simulator-specific workflows.

cs.LG↗