Engineering · 8 min read · April 23, 2026
Automated SysML generation bridges text to engineering models
Hendricks and Cicirello propose a five-step pipeline using NLP and LLMs to convert unstructured documents into SysML diagrams and executable dynamical system models.
Researchers automate conversion of natural language documents into SysML diagrams and computational models for engineering systems.
- — Five-step pipeline extracts domain knowledge from text corpora and system descriptions.
- — SysML diagrams capture component dependencies, attributes, and operations with structured precision.
- — NLP identifies key nouns and relationships; LLMs validate intermediate outputs.
- — Code generation step uses summarization and validation to produce executable models.
- — Domain templates allow integration of expert equations without domain-specific tooling.
- — Pendulum case study shows improvement over raw LLM zero-shot generation.
- — Approach remains agnostic to system type, domain, or target computational software.
Frequently asked
- SysML (Systems Modeling Language) is a formal notation for specifying system architecture, components, and interactions. For dynamical systems, SysML block definition diagrams (BDDs) capture state variables, parameters, and equations in a machine-readable format. This enables automated code generation and reduces ambiguity compared to prose specifications.