The ability to make data-driven decisions has become an essential skill for engineers and many other professionals. To extract meaning from large datasets, data analytics techniques rely on specialized computer software to organize and model the data, identify patterns and draw conclusions.
Working engineers can now catch up on these important skills, which may not have been offered as part of their original training, in an online class called Fundamentals of Industrial Data Analytics offered by the University of Wisconsin-Madison.
Kaibo Liu, an assistant professor of industrial and systems engineering, has developed that class and is teaching it, for the first time, in the spring 2018 semester to students who are enrolled in one of several online master of engineering degree programs offered by the Department of Engineering Professional Development.
The inaugural class (ISyE 412) has attracted 32 working engineers. A more mature audience than Liu’s typical crowd of undergraduates, they range in age from the late 20s to the mid-50s and live in 16 different states and three other countries (Canada, Nigeria and United Arab Emirates). Some of the companies they work for include 3M, Harley-Davidson, Intel, Siemens, Boeing and John Deere.
“The students are self-motivated and eager to learn because they already know how useful data analytics skills will be for their companies,” Liu says. “But they aren’t the only learners here. Many of the questions and problems they bring to the class will be new to me and some may even kick off long-term academic-industry collaborations for our department.”
The class meets in live 60-minute web conferences twice a week. Each week, Liu prepares and edits approximately 60 minutes of video lecture material offline and expects the students to have watched most of those lectures prior to the web conference.
During the live session with Liu, students ask questions about the video material or their homework assignments. To answer these questions, Liu can share his computer screen with his virtual audience to explain the material in more depth.
He calls this style of teaching a hybrid between traditional and flipped classroom teaching. In the latter, the instructor typically prepares online video lectures that students watch at their convenience while they tackle hands-on problems, with input and feedback from the professor, during the scheduled in-person class time.
The second weekly web conference, which doesn’t rely on pre-recorded material, is the lab portion of the class. Here, Liu’s teaching assistant, graduate student Wenjun Zhu, answers questions and walks the students through hands-on exercises designed to teach the software package Tableau and the programming language R. Like the video lectures, the lab sessions are recorded so that students who were unable to attend can review the material online, on their own time.
Liu says teaching an online class for the first time requires much more preparation time than a traditional course (which he is also teaching, in person, in spring 2018 to 55 on-campus students, mostly undergraduates). In addition to the video recording and editing, he has to compress the material he typically covers in two 75-minute lectures into one 60-minute lecture that he then splits into three more digestible 20-minute segments.
But for future classes, Liu expects a reduced workload because he can reuse most of the recorded lectures, meaning he can focus more on the live Q&A sessions and on supervising a class project that involves analyzing and interpreting real datasets.
For the class project, which makes up 25 percent of the final course grade, the students will work in pairs, using either a real dataset from their day job or one provided by Liu. Toward the end of the semester, students will share their results in individual class presentations during the weekly live sessions.
“Online teaching takes some getting used to, for sure,” Liu says. “I think one of the biggest challenges is to maintain student interest and engagement throughout the semester without any in-person interactions. But I’m really excited to teach this class because it’s the first time that the course material I have developed is having an immediate impact well beyond the physical boundaries of the campus.”
Author: Silke Schmidt