Our lab is committed to maximizing the broader impacts of our research through open-source software, educational innovation, workforce development, and community engagement. Below is a summary of our ongoing efforts.
-
Foam-Agent: an end-to-end, composable multi-agent framework for automating CFD simulations in OpenFOAM.
(NeurIPS 2025 Workshop)
-
Deep Autoencoder with SVD Convergence: learnable weighted dimensionality reduction frameworks for fluid dynamics.
- DiagramBank: a dataset of diagram design exemplars with paper metadata for retrieval-augmented generation.
- FoamGPT: dataset and tools for LLM-based computational fluid dynamics.
-
pykoopman: a Python package for data-driven approximation of the Koopman operator.
Published in the Journal of Open Source Software (2023).
-
Neural Implicit Flow (NIF): a library for mesh-agnostic dimensionality reduction of spatiotemporal PDEs.
- Featured Lecture: an open-access lecture on Aerodynamics: Thin Airfoil Theory. More lectures on scientific machine learning, numerical methods, and aerodynamics are available on our YouTube Channel.
- Quasi-1D Nozzle Flow Simulator: an interactive web app for simulating compressible nozzle flows, designed for aerodynamics education. [Live Demo]
- Unsolicited Advice for Ph.D. Students: a publicly available collection of advice on research methodology, academic writing, and career development. Also available on GitHub.
-
RPI Science and Technology Entry Program (STEP) Workshop, April 2025
Prof. Pan led the "AI Meets Engineering" workshop, introducing K-12 students to Generative AI, data-driven modeling, and future career paths in STEM.
Future Career Paths: Engaging families in STEM