Building the Next Generation of Physics‑Informed Models: A Deep Dive into PINN
Learn how physics‑informed neural networks (PINNs) let you build fast, data‑efficient models by embedding physics directly into training. We focus on simple, powerful 1D problems so you can experiment quickly and really understand what’s going on.
Why this webinar
Traditional ML needs lots of labeled data and often ignores known physics. Classical simulation is accurate but rigid, slow, and hard to scale across variants. PINNs bridge this gap by encoding governing equations, boundary conditions, and data into a single learning problem.
We intentionally stay in 1D for this session: systems are easy to visualize, quick to train, and perfect for building intuition around collocation, residuals, and stability—without getting lost in mesh complexity.
Who should attend
- • ML / data scientists working with physical systems
- • Simulation / CAE engineers curious about PINNs
- • R&D and innovation teams exploring physics‑AI
You’ll walk away with
- • Mental model of how PINNs actually work
- • Clear next steps to go deeper via the training course
Agenda (40 minutes)
Part I — Foundations
- • Opening & context (3’)
- • Why physics‑informed models (8’)
- • Anatomy of a modern PINN (8’)
Part II - 1D and 3D in Practice
- • 1D heat conduction
- • Mass‑spring‑damper dynamics
- • Diffusion and beam deflection
Part III — Platform & Next Steps
- • Introducing the oorja PINN platform
- • How to keep experimenting after the session
- • Q&A and training course overview
Speakers
Vineet Dravid, PhD
Founder & CEO, oorja. Background in engineering simulation and AI‑driven product development across batteries and complex systems.
Prashant Srivastava, PhD
Co-Founder and CTO. Head of Research, oorja. PhD (Computational Mechanics), 10+ yrs in multiphysics modeling. Built production PINN pipelines for batteries, thermal systems, and structures.
Register for the webinar
Fill in your details to receive the joining link, calendar invite, and example notebooks after the session.