Starting today
Physics-Informed Neural Networks (PINNs): AI-Driven Scientific Computing & Engineering Simulation
Hands-On Training in Python, Deep Learning, PDE Solvers, Heat Transfer, Fluid Dynamics & Inverse Modelling
About This Course
Physics-Informed Neural Networks (PINNs) are an emerging approach in Scientific Machine Learning that combines artificial intelligence, deep learning and physical governing equations to solve complex scientific and engineering problems.
This international live hands-on workshop provides practical training in developing physics-aware AI models for heat transfer, fluid dynamics and engineering parameter estimation. Participants will learn to integrate physical laws, differential equations and scientific datasets into neural networks using Python, PyTorch, DeepXDE and Google Colab.
Through guided computational exercises, participants will build PINN models, simulate physical systems, evaluate predictions and explore real-world applications in mechanical engineering, computational physics, materials science, energy systems and other engineering disciplines.
The workshop includes a Day 0 preparatory session covering Python setup, dataset downloads and scientific computing fundamentals, making it accessible to participants from multidisciplinary backgrounds.
Aim
To equip students, researchers, academicians and industry professionals with practical knowledge and computational skills in Physics-Informed Neural Networks (PINNs) for AI-driven scientific modelling, engineering simulation and physical parameter estimation using Python-based deep learning frameworks.
Workshop Objectives
1) Understand Physics-Informed AI: Explore the principles of PINNs and their integration with scientific machine learning and physical governing equations.
2) Develop Computational Skills: Gain practical experience with Python, PyTorch, DeepXDE and Google Colab for physics-aware AI modelling.
3) Implement Neural PDE Solvers: Learn to incorporate ordinary and partial differential equations, initial conditions and boundary conditions into neural networks.
4) Simulate Engineering Systems: Develop PINN models for heat transfer and simplified fluid dynamics problems using scientific datasets.
5) Perform Inverse Modelling: Apply PINNs to estimate unknown physical parameters from simulated sensor observations.
6) Evaluate Model Performance: Analyse model accuracy, training convergence and prediction errors through scientific visualization and numerical comparisons.
Workshop Structure
Day 0: Preparatory Session – Python Setup, Dataset Download & Scientific Computing Basics
- Introduction to Physics-Informed Neural Networks (PINNs) and workshop applications.
- Overview of heat transfer, fluid dynamics and inverse modelling case studies.
- Introduction to Python programming and Google Colab for scientific computing.
- Setting up PyTorch, DeepXDE, NumPy, SciPy and Matplotlib.
- Understanding scientific datasets: spatial coordinates, time, temperature, velocity and physical parameters.
- Introduction to heat-equation, Burgers-equation and synthetic sensor datasets.
- Downloading and organizing the required datasets and Google Colab notebooks.
- Loading CSV/NPZ datasets using NumPy and visualizing sample temperature data.
- Testing automatic differentiation and confirming the computational environment.
Hands-On: Set up Google Colab, download the datasets, load sample scientific data and generate a basic temperature visualization.
Input: Prepared heat-transfer dataset, Burgers-equation reference dataset, synthetic sensor data and starter Python notebooks.
Tools Used: Google Colab, Python, NumPy, Matplotlib, PyTorch, DeepXDE.
Output: Ready-to-use Python environment, downloaded datasets and a sample temperature plot.
Day 1: Physics-Informed Neural Networks & AI-Based Heat Transfer Simulation
- Introduction to scientific machine learning and physics-informed deep learning.
- Understanding how physical laws and experimental data are integrated into neural networks.
- Fundamentals of ordinary and partial differential equations (ODEs/PDEs).
- Introduction to governing equations, boundary conditions and initial conditions.
- Neural network architecture, automatic differentiation and physics-informed loss functions.
- Understanding the one-dimensional transient heat equation and thermal diffusivity.
- Preparing spatial-temporal input data for temperature prediction.
- Building a Physics-Informed Neural Network using Python and DeepXDE.
- Training the PINN to predict temperature distributions across space and time.
- Visualizing temperature profiles and analysing model convergence.
- Comparing PINN predictions with analytical or numerical reference solutions.
Hands-On: Develop and train a 1D heat-transfer PINN using DeepXDE, visualize temperature distributions and evaluate prediction accuracy.
Input: 1D heat-equation reference solution, spatial and temporal coordinates, initial/boundary conditions and thermal diffusivity.
Tools Used: Python, Google Colab, PyTorch, DeepXDE, NumPy, Matplotlib.
Output: Trained heat-transfer PINN, temperature heatmap, training-loss graph and predicted-versus-reference temperature plots.
Day 2: AI-Driven Fluid Dynamics, Neural PDE Solvers & Model Validation
- Introduction to computational fluid dynamics (CFD) and physics-informed simulation.
- Understanding nonlinear transport phenomena and their engineering applications.
- Introduction to the one-dimensional viscous Burgers’ equation.
- Understanding advection, diffusion, viscosity and their effects on fluid transport.
- Preparing reference datasets containing spatial coordinates, time and velocity.
- Defining governing equations, initial conditions and boundary conditions.
- Understanding collocation points and physics-based residual minimization.
- Developing a PINN for the Burgers’ equation using DeepXDE.
- Training the neural network to predict velocity distributions.
- Visualizing velocity fields and spatiotemporal flow behaviour.
- Evaluating model predictions using RMSE and relative L2 error.
- Comparing PINN predictions with numerical solutions and identifying model limitations.
Hands-On: Build and train a PINN to solve the 1D Burgers’ equation, predict velocity distributions and compare results with numerical reference data.
Input: Burgers-equation benchmark dataset, spatial-temporal coordinates, viscosity parameter and initial/boundary conditions.
Tools Used: Python, Google Colab, PyTorch, DeepXDE, NumPy, SciPy, Matplotlib.
Output: Trained fluid-transport PINN, velocity heatmap, predicted-versus-reference velocity plots and model error analysis.
Day 3: Inverse Physics-Informed Modelling, Parameter Estimation & Engineering Applications
- Introduction to forward and inverse problems in scientific machine learning.
- Understanding the identification of unknown physical parameters from limited observations.
- Applications of inverse PINNs in thermal systems, structural mechanics, materials and energy engineering.
- Introduction to sparse sensor measurements and physics-guided parameter estimation.
- Preparing synthetic temperature measurements for inverse modelling.
- Understanding thermal diffusivity as an unknown trainable physical parameter.
- Integrating observational data with governing equations in a PINN.
- Developing an inverse PINN using Python and DeepXDE.
- Training the model to estimate thermal diffusivity from temperature observations.
- Comparing estimated physical parameters with known reference values.
- Analysing the influence of sparse observations and measurement noise.
- Visualizing observed-versus-predicted temperature distributions and parameter convergence.
- Preparing a mini-project report with prediction accuracy, estimated parameters and engineering interpretation.
Hands-On: Develop an inverse PINN to estimate unknown thermal diffusivity from synthetic temperature sensor measurements and validate the estimated parameter.
Input: Synthetic temperature sensor dataset, spatial-temporal coordinates, known heat-equation physics and initial/boundary conditions.
Tools Used: Python, Google Colab, PyTorch, DeepXDE, NumPy, SciPy, Matplotlib.
Output: Trained inverse PINN, estimated thermal diffusivity, parameter convergence plot, observed-versus-predicted temperature graph and engineering mini-project report.
Important Dates
Registration Ends
December 2, 2026
IST 4:30 PM
Workshop Dates
December 8, 2026 – December 11, 2026
IST 8:30 PM
Workshop Outcomes
Upon successful completion of the workshop, participants will be able to:
-
Build and train Physics-Informed Neural Networks using Python and DeepXDE.
-
Integrate physical laws and differential equations into deep learning models.
-
Simulate temperature distributions and nonlinear fluid transport using PINNs.
-
Apply inverse modelling techniques for engineering parameter estimation.
-
Validate PINN predictions using numerical reference solutions and error metrics.
-
Develop reproducible scientific computing workflows for interdisciplinary engineering research.
Fee Structure
Ph.D. Scholar / Researcher
₹3,499 | $80
Enrol as Ph.D. Scholar / ResearcherPlus members pay ₹2,799 / $64 · Join Plus
Academician / Faculty
₹4,499 | $90
Enrol as Academician / FacultyPlus members pay ₹3,599 / $72 · Join Plus
Industry Professional
₹5,499 | $105
Enrol as Industry ProfessionalPlus members pay ₹4,399 / $84 · Join Plus
Frequently asked questions
Will I get the recording?
Yes. Paid participants get the session recordings and study material in their NanoSchool dashboard (My registrations → Recorded Lectures / Study Material).
What if I miss a live session?
Watch the recording in your dashboard and send your questions to the mentor through the support desk.
Do I get a certificate?
Yes, an e-certificate of participation after the workshop, with a QR code anyone can verify and an "Add to LinkedIn" button.
Which fee category applies to me?
Choose the category that matches your current role: Student, Ph.D. Scholar / Researcher, Academician / Faculty or Industry Professional.
How can I pay?
By card or PayPal on the payment page, or by UPI when paying in INR; international bank transfer is also available there.
What is the refund policy?
See our Returns, Refunds & Cancellation policy: Returns, Refunds & Cancellation policy
Registration open
More upcoming workshops
Live, mentor-led and certified. Dates and times in IST.
Satellite Remote Sensing & GIS for Seagrass Mapping and Coastal Conservation
- Dates
- 21–23 Nov 2026
- Time
- 5:30 PM IST
- Duration
- 3 Days
- Level
- Moderate
- Student₹2,499
- Ph.D. Scholar / Researcher₹3,499
- Academician / Faculty₹4,499
- Industry Professional₹6,499
Digital Twins: Predictive Modeling for Dynamic Industrial Processes
- Dates
- 25–27 Nov 2026
- Time
- 5:30 PM IST
- Duration
- 3 Days
- Level
- Moderate
- Student₹2,499
- Ph.D. Scholar / Researcher₹3,499
- Academician / Faculty₹4,499
- Industry Professional₹6,499
Applied Survival Analysis in R: Kaplan–Meier Curves, Cox Models & Competing Risks
- Dates
- 25–27 Nov 2026
- Time
- 5:00 PM IST
- Duration
- 3 days
- Level
- Moderate
- Student₹2,199
- Ph.D. Scholar / Researcher₹3,199
- Academician / Faculty₹4,199
- Industry Professional₹5,199
Next-Gen LLM Engineering: Build, Fine-Tune & Deploy Production-Ready AI Applications
- Dates
- 25–27 Nov 2026
- Time
- 5:30 PM IST
- Duration
- 3 Days
- Level
- Moderate
- Student₹2,999
- Ph.D. Scholar / Researcher₹3,999
- Academician / Faculty₹4,999
- Industry Professional₹5,999
Environmental Exposomics in R: PFAS, Chemical Mixtures and Metabolomics
- Dates
- 26–28 Nov 2026
- Time
- 5:30 PM IST
- Duration
- 3 Days
- Level
- Moderate
- Student₹2,499
- Ph.D. Scholar / Researcher₹3,499
- Academician / Faculty₹4,499
- Industry Professional₹6,499
AI Cloud Defender: AWS Threat Detection & Explainable SOC Analytics
- Dates
- 27–29 Nov 2026
- Time
- 5:00 PM IST
- Duration
- 3 days
- Level
- Moderate
- Student₹2,299
- Ph.D. Scholar / Researcher₹2,999
- Academician / Faculty₹3,699
- Industry Professional₹4,500






