Searcharxiv⌕ Search

arXiv subjects

Ubaldo M. Córdova-Figueroa

Publications and source records attributed to Ubaldo M. Córdova-Figueroa.

3 recordsLinked to original sources

DynaMate2: runtime registration of expert-defined tools for agentic scientific workflow automation

Agentic large-language-model systems can coordinate scientific tools, but many implementations remain difficult for domain scientists to extend without modifying the source orchestration code or relying on unconstrained code generation. DynaMate2 is a LangGraph-based multi-agent framework for converting expert-defined Python functions into persistent AI-callable tools. The architecture separates domain execution from LLM supervision: registered tools perform scientific operations, while a supervisor LLM decomposes goals, selects specialist agents, routes inputs, and propagates outputs across steps. DynaMate2 supports: runtime tool registration from inline code, source files, and explicitly requested natural-language specifications; persistent storage of tools, agents, and conversation state; and a web interface for interactive workflow assembly. We demonstrate the framework on a molecular simulation workflow in which a single instruction retrieves a MACE foundation model, builds a NaCl-water configuration, runs an ASE molecular dynamics trajectory, and generates energy and temperature diagnostics. The demonstration illustrates how validated workflow components can be composed into a supervised agentic pipeline without rewriting the framework. DynaMate2 therefore provides a reusable template for extending LLM-based automation to research groups with existing Python workflows, while preserving the need for explicit tool validation, reproducibility logs, and deployment-specific safeguards.

physics.chem-ph↗

Dynamics of a Bottom-Heavy Janus Particle Near a Wall Under Shear Flow

After our attention was brought to recent work by Pelargonio and Zaconne regarding the deviation of the white noise function for systems at very large shear, we have included the work as a reference of interest to our readers and clarified the definition of our Peclet number with respect to the Brownian diffusion of the particle as well as included a brief discussion of the magnitude of our shear rate compared to the bath frequency to clarify that the standard white noise applies in our system.

cond-mat.soft↗

Dynamics of a microswimmer near a curved wall: guided and trapped locomotions

We propose a combined analytical-numerical strategy to predict the dynamics and trajectory of a microswimmer next to a curved spherical obstacle. The microswimmer is actuated by a slip velocity on its surface and a uniformly valid solution is provided by utilizing the Reynolds reciprocal theorem in conjunction with the exact hydrodynamic solution of translation/rotation of a sphere in an arbitrary direction next to a stationary obstacle. This approach permits the hydrodynamic interaction of the microswimmer and the obstacle to be consistently and accurately calculated in both far and near fields. Based on the analysis, it was shown that while the "point-singularity solution" is valid when the microswimmer is far from the obstacle, it fails to predict the correct dynamics when the swimmer is close to the obstacle (i.e. gap size is approximately twice the characteristic length of the microswimmer). Two different paradigms for propulsion, so-called "squirmer" and "phoretic" models, were examined and for each type of microswimmer, various types of trajectories are highlighted and discussed under which circumstances the swimmer can be hydrodynamically trapped or guided by the obstacle. The analysis indicates that it is always easier to capture a microswimmer in a closed circular orbit with a large sized obstacle ($\sim 20$-$40$ times larger than the microswimmer size) as in this case the magnitude of rotational velocity can be sufficiently large that the swimmer can adjust its distance and orientation vector with the obstacle.

cond-mat.soft↗