These lecture materials are openly available to everyone.
For students: You are encouraged to use these materials to support your studies.
For instructors: You are welcome to use, modify, and distribute these materials in your teaching.
No credit or reference to us is required.
This course develops a unified framework for integrating physics-based modeling, machine learning, control, and embedded deployment in engineering systems. The emphasis is on closing the loop from modeling to real-world execution under physical constraints. Physical AI is not only about learning models. It is about building systems that operate reliably in the real world.
Learning Objectives
Formulate learning problems with physical structure and constraints
Develop hybrid physics + data-driven models
Design control systems with learning components
Implement real-time AI on embedded hardware
Deploy models from simulation to real-world systems
Topics HTML Colab Slides PowerPoints PS Solution
[Part I: Foundations of Physical AI]
Introduction to Physical AI
Dynamical Systems and Modeling
Probabilistic Modeling and Estimation
[Part II: Physics-Informed Learning]
Physics-Informed Neural Networks (PINNs)
Neural Operators
Constraints and Structure in Learning
[Part III: Learning and Control]
Reinforcement Learning for Physical Systems
Physics-Guided and Safe RL
[Part IV: Sim-to-Real Transfer]
Sim-to-Real Gap and Domain Randomization
System Identification and Residual Learning
[Part V: Embedded Physical AI (Hands-on)]
Embedded Systems and Real-Time Control
Edge AI and AI-in-the-Loop Control
Sim-to-Real Deployment on Hardware
System Integration & Demo