University of Michigan mechanical engineering faculty members use artificial intelligence to advance educational experiences for students, better preparing them for post-grad opportunities
Artificial Intelligence (AI) is revolutionizing the field of engineering by transforming how problems are analyzed, solutions are designed, and systems are optimized across a wide range of disciplines. Several U-M ME faculty members have begun to integrate AI and machine learning techniques into their coursework to prepare students for careers in a future where AI technologies are becoming increasingly important.
Chenhui Shao, associate professor of mechanical engineering, created a new course, Data Science for Manufacturing Quality Control (ME 401), to help expose students to data science concepts crucial in manufacturing careers. In this course, students are tasked with working with large-scale, real-world data sets to find and solve problems associated with quality control.
“This course aims to teach students the fundamentals of quality management in the big data era,” Shao said. “It also teaches them to use state-of-the-art data science methods including machine learning, AI, and statistics to solve manufacturing factory floor quality problems.”
The manufacturing industry has experienced a recent shift with the advent of smart manufacturing, which uses data derived from machines and sensors to optimize and improve the production process. Smart manufacturing allows companies to automatically adjust to changes in demand and business needs while maintaining the quality of their products.
“At most universities, we don’t teach something like an AI or machine learning course tailored to manufacturing or engineering,” Shao said. “I feel that this is a must-have course. So really, the goal is to fill this curriculum gap so we can attract them to manufacturing, but also to give them the skills they need to adapt to the profession.”
Manufacturing data is unique and cannot rely on methodologies developed for other non-manufacturing applications, Shao explained. Instead, approaches to manufacturing data management must be tailored specifically to industrial applications. Teaching students to use AI for manufacturing-specific purposes is a key goal of ME 401.
The course is designed to present students with problems and encourage them to formulate solutions using the data they are given, Shao said. Throughout the semester, students complete labs on topics including data analysis, measurement system analysis, feature engineering, and AI model development. These labs, combined with homework assignments, teach students how to use machine learning techniques to develop effective strategies for quality control.
Despite its benefits to the manufacturing industry, smart manufacturing has also revealed a growing gap between the skill set of the current workforce, which is largely uneducated in AI technologies, and the demand for data-driven decision making. Compounded by the fact that the U.S. manufacturing industry is expected to grapple with 2.1 million unfilled jobs by 2030, there is a need for growth in the workforce.
ME 401 aims to bridge the gap by providing current college students with the necessary tools to excel in manufacturing careers in the age of machine learning after graduation.
Wei Lu, professor of mechanical engineering, is leading a broad effort to equip mechanical engineers with the knowledge and skills to harness the power of AI and machine learning. Lu’s approach, however, is open to the public through Coursera, an online learning platform that partners with leading organizations to provide free and low-cost online educational opportunities.
Lu recently published a series of three Massive Open Online Courses (MOOCs) titled, “AI for Mechanical Engineers,” which is designed to help mechanical engineers learn how using AI can benefit their work.
“From optimizing design processes to revolutionizing autonomous systems and addressing energy and biomedical challenges, they will delve into cutting-edge AI methodologies tailored for mechanical engineering contexts,” Lu said. “After completing the series, they will be equipped with the knowledge and skills to navigate the evolving landscape of mechanical engineering and drive transformative change through the integration of AI technologies.”
Lu’s MOOCs consist of a series of modules and exercises that encourage students to develop their AI knowledge in practical, real-world contexts. This approach encourages students to think both critically and creatively, allowing them to adapt their knowledge to dynamic situations while seamlessly connecting theoretical concepts to their practical implementation.
Students in Lu’s courses will be exposed to a real engineering problem—such as optimizing a product design, predicting battery health, or enabling a self-driving car to make decisions—to help them understand why AI is useful before learning how it works. Students are then tasked with using tools like Python and Pytorch to explore, build, and train machine learning models themselves.
“This approach builds motivation and context, making even complex topics feel purposeful and approachable,” Lu said.
Presenting the content in a series offers additional benefits to students by creating a workflow in which they can develop and apply their skills. The first course, “AI for Design and Optimization,” exposes students to fundamental concepts of AI for engineering and design optimization processes.
The other two courses, “AI for Autonomous Vehicles and Robotics,” and “AI for Energy and Biomedical Applications,” allow students the opportunity to apply those fundamentals to specialized areas.
Lu said that even beyond individual skill-building, his goal is to help shape a new generation of engineers who can merge traditional engineering principles with artificial intelligence and data science.
The growth of AI and Machine Learning in the mechanical engineering discipline has led to a shift in how engineers approach and solve problems. This development has made it necessary for current and future engineers to learn valuable AI principles.
“In a world increasingly driven by data and intelligent systems, understanding and applying AI and machine learning techniques is becoming essential for mechanical engineers who aim to stay competitive, innovative, and impactful,” Lu explained.
