Introducing AI into an undergraduate Kinematics of Machines course

Heather Louise Lai

Download Paper | Permalink

Conference: 2024 Fall ASEE Mid-Atlantic Section Conference
Location: Farmingdale State College, NY, New York
Publication Date: October 25, 2024
Start Date: October 25, 2024
End Date: November 5, 2024
Conference Session: Technical Sessions 1
Tagged Topics: Diversity and Professional Papers
Page Count: 7
DOI: 10.18260/1-2--49449
Permanent URL: https://peer.asee.org/49449
Download Count: 16

Paper Authors

biography

Heather Louise Lai State University of New York at New Paltz orcid.org/0000-0002-9618-2739

visit author page

Heather Lai is an Associate Professor of Mechanical Engineering at SUNY New Paltz, NY where she teaches courses in dynamics, system dynamics, finite element analysis and computer simulation. Her professional background and research interests include automotive vibration (Motorola Inc.), musculoskeletal biomechanics (BME, Wayne State University), room acoustics, wind farm acoustics and the dynamic behavior of 3D printed multi-materials. Over the past 8 years, she has collaboratively developed a number of new and revised courses, including a new System Dynamics Lab. She has also worked with a number of SUNY students to investigate different aspects of 3D printed multi-material structures.

visit author page

Download Paper | Permalink

Abstract

A first attempt at adding machine learning / AI into a 3rd year mechanical engineering Kinematics of machines course is discussed. The course, required for all mechanical engineering students, involves the kinematic analysis of linkages, gears and cams following a traditional course outline. The course has been modified to include more computer-related solution techniques, without increasing the required prerequisite knowledge of the students. The portions of the course which are most significantly affected by this change are the sections related to analysis and syntheses of 4 bar mechanisms. Several topics including graphical methods for performing position, velocity and acceleration analysis of linkage have been truncated to make room for content related to computational computer-based methods. The coding language used for the computer programming in this course is MATLAB. As computer programming is not a prerequisite for the course, the course scaffolding must include the basic structure for all the necessary programming code. A kinematics of machines textbook with extensive MATLAB code has been selected in order to provide guidance to the students. In addition to this, prebuilt sections of code will be made available, providing building blocks for the students to assemble and tailor their code to the specific characteristics of the linkage they are analyzing and the goals of their computations. Vector loop analysis is used to develop position, velocity and acceleration analytical equations. Solution techniques for the vector loops will include traditional hand calculation based methods as well as several computer based computational methods. Code for performing Newtons’ method is taught along with the use of more complex built in MATLAB functions such as fzero and the solve command. Optimization techniques for mechanism synthesis will also be added to the course, providing students with an opportunity to go beyond traditional trial and error methods and utilize MATLAB’s optimization toolbox to improve on their designs. The use of machine learning in mechanism optimization will also be covered and used for simple examples. The extent that it can be used for more general problems is limited by access to algorithms and computational resources. This is a work in progress, to be presented during the semester of the first implementation. The instructor’s attempts will be described in order to prompt discussion related to the implementation of AI in traditional engineering courses.

Citation
Format

Lai, H. L. (2024, October), Introducing AI into an undergraduate Kinematics of Machines course Paper presented at 2024 Fall ASEE Mid-Atlantic Section Conference, Farmingdale State College, NY, New York. 10.18260/1-2--49449

TY - CPAPER
AB - A first attempt at adding machine learning / AI into a 3rd year mechanical engineering Kinematics of machines course is discussed. The course, required for all mechanical engineering students, involves the kinematic analysis of linkages, gears and cams following a traditional course outline. The course has been modified to include more computer-related solution techniques, without increasing the required prerequisite knowledge of the students.
The portions of the course which are most significantly affected by this change are the sections related to analysis and syntheses of 4 bar mechanisms. Several topics including graphical methods for performing position, velocity and acceleration analysis of linkage have been truncated to make room for content related to computational computer-based methods.
The coding language used for the computer programming in this course is MATLAB. As computer programming is not a prerequisite for the course, the course scaffolding must include the basic structure for all the necessary programming code. A kinematics of machines textbook with extensive MATLAB code has been selected in order to provide guidance to the students. In addition to this, prebuilt sections of code will be made available, providing building blocks for the students to assemble and tailor their code to the specific characteristics of the linkage they are analyzing and the goals of their computations.
Vector loop analysis is used to develop position, velocity and acceleration analytical equations. Solution techniques for the vector loops will include traditional hand calculation based methods as well as several computer based computational methods. Code for performing Newtons’ method is taught along with the use of more complex built in MATLAB functions such as fzero and the solve command. Optimization techniques for mechanism synthesis will also be added to the course, providing students with an opportunity to go beyond traditional trial and error methods and utilize MATLAB’s optimization toolbox to improve on their designs. The use of machine learning in mechanism optimization will also be covered and used for simple examples. The extent that it can be used for more general problems is limited by access to algorithms and computational resources.
This is a work in progress, to be presented during the semester of the first implementation. The instructor’s attempts will be described in order to prompt discussion related to the implementation of AI in traditional engineering courses.

AU - Heather Louise Lai
CY - Farmingdale State College, NY, New York
DA - 2024/10/25
PB - ASEE Conferences
TI - Introducing AI into an undergraduate Kinematics of Machines course
UR - https://peer.asee.org/49449
DO - 10.18260/1-2--49449
ER -