1 of 27

AI for Solving Mechanical Engineering Problems

Prof. Seungchul Lee

Industrial AI Lab.

2 of 27

Course Information For ME491

  • Course title: AI for Solving Mechanical Engineering Problems

​

  • Instructor: Prof. Seungchul Lee
    • Office: N7-4, 6102
    • Email: seunglee@kaist.ac.kr

​

  • Students are anticipated to acquire knowledge of machine learning and deep learning algorithms used in data analytics and their practical implementations in Python. Although mathematical techniques and theoretical aspects will be included, the principal objective is to furnish students with the skills and fundamental principles essential for addressing data-related challenges encountered within the realm of mechanical engineering. The majority of the illustrative instances will be directly relevant to mechanical engineering.

​

3 of 27

Lecture Materials

4 of 27

Artificial Intelligence

5 of 27

Mechanical Engineering and AI

  • AI for mechanical engineering
    • New tool to better understand physical phenomena or discover better designs

​

  • To bring/develop the state-of-the-art AI technologies for mechanical engineering problem

​

​

6 of 27

Q2) Prerequisites

7 of 27

Q2) Prerequisites

​

​

​

​

​

​

​

​

​

​

​

  • Recommended Prerequisite
    • ME492 Programming for Solving Mechanical Engineering Problems

8 of 27

Q3) Topics and Scope

9 of 27

Course Schedule

Week

Contents

Assignments

Week 1

Introduction to Data-driven Approach

​

Week 2

Mathematics for AI: Linear Algebra

HW#01

Week 3

Mathematics for AI: Optimization

HW#02

Week 4

Machine Learning: Regression

HW#03

Week 5

Machine Learning: Classification

HW#04

Week 6

Machine Learning: Clustering and Dim. Reduction

HW#05

Week 7

AI in Mechanical Engineering 1

HW#06

Week 8

Midterm

​

Week

Contents

Assignments

Week 9

Deep Learning: ANN, Autoencoder

​

Week 10

Deep Learning: CNN, FCN

HW#07

Week 11

Deep Learning: RNN, LSTM

HW#08

Week 12

Deep Learning: Generative AI

HW#09

Week 13

Physics-informed AI (PINN and DeepONet)

HW#10

Week 14

AI in Mechanical Engineering 2

HW#11

Week 15

AI in Mechanical Engineering 3

​

Week 16

Final Exam and Project

​

10 of 27

Python

  • Python coding example

11 of 27

Linear Algebra

  •  

12 of 27

Optimization

  • Least squares
  • Convex optimization (cvx or cvxpy)
  • Gradient descent

13 of 27

Regression (Data Fitting or Approximation)

  • Statistical process for estimating the relationships among variables

​

14 of 27

Classification

  • The problem of identifying to which of a set of categories (sub-populations) a new observation belongs, on the basis of a training set of data containing observations (or instances) whose category membership is known
  • To find classification boundaries

15 of 27

Dimension Reduction

  • Multiple Sensors + Principal Components
  • the process of reducing the number of random variables under consideration, and can be divided into feature selection and feature extraction.

​

​

16 of 27

Clustering

17 of 27

Deep Artificial Neural Networks

  • Complex/Nonlinear universal function approximator
    • Linearly connected networks
    • Simple nonlinear neurons

​

​

​

Class 1

Class 2

…

…

…

…

…

…

Output

Input

18 of 27

Topics

From deeplearning.mit.edu

https://github.com/lexfridman/mit-deep-learning/blob/master/tutorial_deep_learning_basics/deep_learning_basics.ipynb

19 of 27

Double Pendulum (Dynamics)

SEO, Sungyong, et al. Controlling neural networks with rule representations. Advances in neural information processing systems, 2021.

20 of 27

Euler Beam (Solid Mechanics)

 

 

 

 

21 of 27

Fluid Mechanics

22 of 27

Q4) How To Study

23 of 27

Jupyter Notebook

  • Jupyter notebook and Colab

24 of 27

All lecture materials are already available at

​

Machine Learning

Deep Learning

25 of 27

Assignments

  • Homework Assignments
    • Hand-written and programming exercises will be posted.
    • They are due one week after they are posted.

​

  • Midterm and Final Exams
    • There will be two parts: hand-written exam and coding.

​

  • Projects
    • The goal of the final project is to provide you with insight into what an AI research project entails, and ideally, spark your enthusiasm for engaging in research within the mechanical engineering field. We aim to provide an opportunity to gain in-depth experience developing an AI-based approach in solving mechanical engineering problems.�

26 of 27

Grading

  • Homework 20 %

​

  • Midterm 40 %
  • Final 30 %

​

  • Project 10%

​

27 of 27

Coding

  • Python in class, assignments, and project
    • Used a lot
    • Lots of coding problems in both homework and exam

​

  • Projects
    • The goal of the final project is to provide you with insight into what an AI research project entails, and ideally, spark your enthusiasm for engaging in research within the mechanical engineering field. We aim to provide an opportunity to gain in-depth experience developing an AI-based approach in solving mechanical engineering problems.

​