Concerning the Integration of Machine Learning Content in Mechatronics Curricula

Concerning the Integration of Machine Learning Content in Mechatronics Curricula

Jörg Frochte (Bochum University of Applied Sciences, Germany), Markus Lemmen (Bochum University of Applied Sciences, Germany) and Marco Schmidt (Bochum University of Applied Sciences, Germany)
Copyright: © 2020 |Pages: 22
DOI: 10.4018/978-1-5225-7793-5.ch004

Abstract

Machine learning is becoming more and more important for mechatronic systems and will become an ordinary part of today's student life. Thus, it is obvious that machine learning should be part of today's student's curriculum. Unfortunately, machine learning seldomly is implemented into the curriculum in a substantial or linking manner, but rather offered as an elective course. This chapter provides an analysis of how machine learning can be integrated as a mandatory part of the curriculum of mechatronic degree courses. It is considered what the required minimal changes in fundamental courses should be and how traditional subjects like robotics, automation, and automotive engineering can profit most of this approach. As a case study, this chapter utilizes an existing German mechatronic degree course specialized on information technology, which covers most of the discussed aspects.
Chapter Preview
Top

An Overview On Machine Learning

Which aspects of machine learning are now the most important ones for mechatronics? One taxonomy for machine learning approach divides them into

  • Supervised,

  • Unsupervised

  • And reinforcement methods.

Key Terms in this Chapter

Machine Learning: A branch of artificial intelligence. Summarizes different approaches to “learn” from data.

Mechatronic Degree Courses: A complete mechatronics course at the university with a total length of 6 to 8 semesters. The student completes the course with a bachelor thesis.

Unsupervised Learning: Machine learning approach used for clustering. Learns from unlabeled data (i.e., there is no supervision from a human who specifies the labels for the data).

Supervised Learning: Machine learning approaches often used for regression and classification.

Reinforcement Learning: Machine learning approaches often used in robotics. A reward is used to teach a system a desired behavior.

Complete Chapter List

Search this Book:
Reset