(Updated on Feb 05, 2024 by Dr. Song Chaoyang)

Summary of Teaching at BionicDL

Teaching is an essential part of my profession, not just to deliver knowledge and skills to the students but also a way of self-reflection between textbook understanding and emerging development in academia and industry, as well as an opportunity to learn from our students. I have applied this philosophy since my first Lecturer appointment at Monash University, which I carried on to my current position as an Assistant Professor with the Southern University of Science and Technology (SUSTech). In the following, I will summarize my teaching activities at SUSTech in three parts.

Please refer to the Team page for a complete list of students I have supervised since joining SUSTech in June 2018.

I insisted on having all my graduate students, especially during the first semester when they joined SUSTech, participate in my teaching activities, mainly during the lab session, providing me with the opportunity to get to know them and share my way of research and teaching. Another approach I adopted was the joint development of teaching tools and course adjustment with students previously taking my class to join the course development in the next semester, providing me with a fresh understanding of the students’ perspectives on how they would like to design and proceed with their learning experience at SUSTech.

As an Assistant Professor at SUSTech, I fulfilled and exceeded my teaching obligations, opting not to take a probationary teaching exemption for one semester.

I led two courses: ME303/ME311 on Mechanical Design, a compulsory course for all undergraduates in the department and essential for the National Graduate Entrance Exam, and ME336 on Collaborative Robot Learning, mandatory for Robotics Engineering majors.
I engaged in teaching reform efforts, integrating innovative course delivery methods, and was recognized with a 480K teaching grant.
- My involvement extended to international educational collaborations, notably with MIT, to enhance the teaching experience for SUSTech students through the joint centers for mechanical engineering research and education. All my course materials are freely available online, benefiting students within and beyond SUSTech.
In 2023, I collaborated with Professors Dai Jiansheng, Pan Yang, Jia Zhenzhong, and Wang Zheng to create the SUSTech Virtual Teaching Lab for Machine Intelligence Design and Learning, serving as its Director and Principal Investigator.

In December of 2023, I received the UNESCO-ICHEI Higher Education Digitalization Pioneer Case Award for the DeepClaw project, a teaching initiative I developed to facilitate robot learning.

DeepClaw stands out for its innovative approach, affordability, and practicality, making it a valuable educational resource with global accessibility. I ensure that all my graduate students participate in lab sessions, particularly in their first semester, allowing me to share my research and teaching philosophy. Additionally, I co-develop DeepClaw teaching tools with students who have previously taken my courses, providing fresh insights into student learning preferences and enhancing the educational experience at SUSTech.

This summary will end with concluding remarks and a future outlook on my teaching activities at SUSTech.

Teaching Awards

2023	UNESCO-ICHEI Higher Education Digitalization Pioneer Case Award
	The International Centre for Higher Education Innovation under UNESCO (UNESCO-ICHEI)
	To encourage and commend IIOE partner institutions in the world active exploration, innovation, and practice in the field of digitalization of higher education. The award sets up a specific theme every three years, awarded for the first time in 2023, and rewards teams and individuals who have made outstanding contributions to further promoting digital transformation, improving the quality of higher education teaching, and achieving sustainable development, and supports their exploration in promoting higher education innovation.

Teaching Grants

Since joining SUSTech, awarded 480K RMB funding for Teaching Reform Projects as PI.

Grants awarded as PI: University Level: (100K+80K+300K) @ SUSTech

2023.01 – 2024.12	Educational and Teaching Research and Reform Project		PI
	(Grant No.: Y01331838)
	Southern University of Science and Technology Machine Intelligence Design and Learning Virtual Teaching Lab		100K RMB

2021.01 – 2021.12	Educational and Teaching Research and Reform Project		PI
	(Grant No.: Y01331824)
	Southern University of Science and Technology Laboratory Teaching Reform using Cross-Class, Continuous-Project, Big-Team Projects		80K RMB

2019.01 – 2019.12	Educational and Teaching Research and Reform Project		PI
	(Grant No.: 34/K18341702)
	Southern University of Science and Technology Using Collaborative Robots for Interactive Teaching in Emerging Engineering Subjects		300K RMB

2018.12 – 2023.12	SUSTech-MIT Joint Centers for Mechanical Eng. Research & Education		Participant
	(Grant No.: Y01346002)	SUSTech PI: Prof. Fu Chenglong MIT PI: Prof. Harry Asada
	Science, Technology & Innovation Commission of Shenzhen Municipality Interactive Human-Robot System for Robotics Education and Research

Teaching Lab

2023	SUSTech Virtual Teaching Lab for Machine Intelligence Design and Learning
	Director & PI Lab Members: Prof. Dai Jiansheng, Prof. Pan Yang, Prof. Jia Zhenzhong, Prof. Liu Sicong, Prof. Wan Fang, and Prof. Wang Zheng

Teaching Duties

Since joining SUSTech, as the lead instructor of three courses, all accessible online.

ME303/311 Total Enrollment: 238 Total Lecture Hours: 384
ME336 Total Enrollment: 130 Total Lecture Hours: 304
ME49X Total Enrollment: (79) Total Lecture Hours: Not Applicable
Design Lab

ME303 Introduction to Mechanical Design (Lead Instructor in Charge of Course Development):

Autumn 2023 Enrollment: 55 (Coded as ME311, 16 Weeks, 48 Lecture Hours, 3 Credit Units)
Autumn 2022 Enrollment: 27 (16 Weeks, 64 Lecture Hours, 3 Credit Units)
Autumn 2021 Enrollment: 28 (16 Weeks, 64 Lecture Hours, 3 Credit Units)
Autumn 2020 Enrollment: 68 (16 Weeks, 96 Lecture Hours, 3 Credit Units)
Autumn 2019 Enrollment: 38 (16 Weeks, 64 Lecture Hours, 3 Credit Units)
Autumn 2018 Enrollment: 22 (16 Weeks, 48 Lecture Hours, 3 Credit Units)

ME336 Collaborative Robot Learning (Lead Instructor in Charge of Course Development):

Spring 2023 Enrollment: 44 (16 Weeks, 60 Lecture Hours, 3 Credit Units)
Spring 2022 Enrollment: 35 (16 Weeks, 60 Lecture Hours, 3 Credit Units)
Spring 2021 Enrollment: 34 (16 Weeks, 60 Lecture Hours, 3 Credit Units)
Spring 2020 Enrollment: 8 (16 Weeks, 64 Lecture Hours, 3 Credit Units)
Spring 2019 Enrollment: 9 (16 Weeks, 60 Lecture Hours, 3 Credit Units)

ME491 Engineering Practice (Lead Instructor in Charge of Course Development)

Spring 2023 Enrollment: 6 (ME492), 6 (ME495), 6 (ME494), 4 (ME493), 1 (ME491)
Spring 2022 Enrollment: 6 (ME495), 4 (ME493), 1 (ME491), 6 (ME494), 8 (ME492)
Autumn 2021 Enrollment: 1 (ME492)
Spring 2021 Enrollment: 1 (Industrial), 3 (ME494), 6 (ME495), 1 (Social), 5 (ME493), 1 (ME492)
Autumn 2020 Enrollment: 10 (ME492)
Summer 2020 Enrollment: 2 (ME494)
Summer 2019 Enrollment: 1

Innovation Project for Undergraduate Students

Spring 2020 Enrollment: 8

COE492 Capstone Project (Equivalent to Undergraduate Thesis)

Autumn 2023 Enrollment: 3 (Two Semesters at Year 4)
Autumn 2022 Enrollment: 5 (Two Semesters at Year 4)
Autumn 2021 Enrollment: 5 (Two Semesters at Year 4)
Autumn 2020 Enrollment: 6 (Two Semesters at Year 4)
Autumn 2019 Enrollment: 5 (Two Semesters at Year 4)

COE100 Introduction to Engineering

Spring 2023 Enrollment: 38 (1 Lecture)

Supervisees

Since joining SUSTech, 42 students graduated, including 37 Undergraduate and 5 Master Students, serving as their thesis supervisors, among which 20 for Jobs, 5 for Ph.D. study, and 17 for Master study. There are another 18 students currently under my supervision for thesis, including 7 Ph.D., 7 Master’s, and 4 Undergraduate Students.

Research Assistant Professor

2023 – Present

Yi Juan

Ph.D. from The University of Hong Kong

Postdoctoral Scholars

2023 – Present

Xie Kege

Ph.D. from the University of Nottingham, Ningbo, China
(Co-supervised with Prof. Deng Hui @ SUSTech)

Visiting Scholars

2020 – 2022	Tian Hao	Post-Doc visiting from HKU. Joined Huawei
2020 – 2022	Yan Youcan	Doctoral Student visiting from CityU@HK. Joined CERN as a Post-Doc
2021 – Present	Zhong Shuqiao	Doctoral Student visiting from Prof. Lin Jian’s group at SUSTech

Ph.D. Students

2014 – 2014	Timothy Li	Singapore University of Technology and Design (Co-supervised with Prof. Kevin Otto)
2015 – 2017	Chen Yaohui	Monash University (Currently a Faculty Member with Huazhong University of Agriculture. Transferred to Dr. Hoam Chung)
2015 – 2017	Zhang Dongyu	Monash University (Co-supervised with Dr. Hoam Chung)
2019 – Present	Liu Xiaobo	SUSTech Doctoral Student
2019 – Present	Guo Ning	SUSTech Doctoral Student
2019 – Present	Yang Linhan	HKU-SUSTech Joint Doctoral Student Co-supervised with Prof. Pan Jia at the University of Hong Kong
2020 – Present	Gu Yuping	HKU-SUSTech Joint Doctoral Student Co-supervised with Prof. Pan Jia at the University of Hong Kong
2021 – Present	Sun Haoran	HKU-SUSTech Joint Doctoral Candidate Co-supervised with Prof. Pan Jia at the University of Hong Kong
2021 – Present	Wang Ziqian	Pengcheng-SUSTech Joint Doctoral Candidate Co-supervised with Dr. Lu Haibo at Pengcheng Lab
2022 – Present	Han Xudong	SUSTech Doctoral Candidate

Master Students

2013 – 2014	Bangqi Yin	Singapore University of Technology and Design & MIT
2013 – 2014	Aditya Ranjan	Singapore University of Technology and Design & MIT
2016 – 2017	Wu Tong	Monash University
2018 – 2020	Liu Sheng	Joined Electronics Design Institute
2018 – 2020	He Yang	Joined Harmen Karden
2019 – 2022	Wang Zhenhong	Joined Shaoyin Tech.
2020 – 2023	Feng Shihao	Joined RVbust Tech.
2020 – 2023	Guo Yuqin	Joined Hong Kong Public Service Department
2021 – Present	Ge Sheng	Master Class of 2024
2021 – Present	Jie Yu	Master Class of 2024
2021 – Present	Jiang Bingfa	Master Class of 2024
2022 – Present	Dong Yujian	Master Class of 2025
2022 – Present	Xu Ronghan	Master Class of 2025
2023 – Present	Wu Tianyu	Master Class of 2026
2023 – Present	Huang Bangchao	Master Class of 2026

Undergraduate Students

2015 – 2019	Guo Weijie	Joined SUSTech as a Lab Technician
2016 – 2021	Luo Qichen	Joined SUSTech Master Program
2016 – 2021	Ma Zhengjia	Joined Tencent
2016 – 2021	Zheng Shuxin	Joined Huawei
2016 – 2020	Gu Yuping	Joined SUSTech-HKU Jointed Doctoral Program
2016 – 2020	Ge Sheng	Joined SUSTech Master Program
2016 – 2020	He Haibin	Joined The University of Hong Kong (Computer Science Major)
2016 – 2020	Sun Haoran	Joined SUSTech-HKU Jointed Doctoral Program
2016 – 2020	Wu Weipeng	Joined ByteDance (Computer Science Major)
2016 – 2020	Zhang Siyu	Joined SenseTime (Computer Science Major)
2016 – 2020	Feng Shihao	Joined SUSTech Master Program (Non-SUSTech Undergrad)
2017 – 2021	Zhang Zixin	Joined CMU Master Program
2017 – 2021	Jiang Haiyang	Joined DoraBot
2017 – 2021	Jing Yonglin	Joined SUSTech Master Program
2017 – 2021	Wang Baiyue	Joined University of Michigan Master Program
2017 – 2021	Liu Ziqi	Joined Master Program in Hong Kong
2017 – 2021	Chen Mingdong	Joined Brown University Master Program in Industrial Design
2017 – 2021	Xu Tianyuan	Joined Huawei (Computer Science Major)
2017 – 2021	Chai Tianhao	Joined Baidu (Computer Science Major)
2017 – 2021	Shi Jianping	Joined Tencent (Computer Science Major)
2017 – 2021	Wang Zhiwei	Joined Tencent (Computer Science Major)
2018 – 2022	Han Xudong	Joined SUSTech Doctoral Program
2018 – 2022	Xiao Yang	Joined Job Market
2018 – 2022	Qiu Nuofan	Joined SUSTech Master Program
2018 – 2022	Wei Jinqi	Joined DoraBot
2018 – 2022	Dong Yujian	Joined SUSTech Master Program
2018 – 2022	Yang Yuxuan	Joined SUSTech Master Program
2018 – 2022	Xu Ronghan	Joined SUSTech Master Program
2018 – 2022	Yu Zhiyang	Joined SUSTech Master Program (Computer Science Major)
2018 – 2022	Ma Shuyuan	Joined Shaoyin Tech.
2019 – 2023	Yin Jiayi	Joined SUSTech Doctoral Program
2019 – 2023	Huo Jiayu	Joined KTH Master Program
2019 – 2023	Wang Jingran	Joined Columbia University Master Program
2019 – 2023	Tian Feng	Joined ETH Master Program
2019 – 2023	Yin Caihao	Joined Job Market
2019 – 2023	Wu Tianyu	Joined SUSTech Master Program (Non-SUSTech Undergrad)
2020 – 2024	Qiu Wanghongjie	Undergraduate Class of 2024. To Join SUSTech Master Program
2020 – 2024	Zhang Zishang	Undergraduate Class of 2024. To Join SUSTech Master Program
2021 – 2025	Qiu Jianuo	Undergraduate Class of 2025
2021 – 2025	Fan Wangzhuo	Undergraduate Class of 2025

Highlights on Student Learning Outcomes

In this section, I briefly summarize a few highlights of my teaching activities at SUSTech through the students’ learning outcomes.

Teaching Tool Development and International Award

The DeepClaw system went through quite a few iterations of development, initiated by me in 2018 and then passed on to the hands of our students over the years. They were students who first encountered the DeepClaw system when they were taking the ME336 course, then became the designers of the system to optimize it for the next batch of students, practically in charge of their own learning experience with a strong self-motivation, which I felt proud to be part of the transformative journey. In 2023, the DeepClaw Project was selected for the UNESCO-ICHEI Higher Education Digitalisation Pioneer Case Award, which is given for the first time every three years, and the only winning project from SUSTech, evaluated on a global scale.

Cloud-based CAD System and Lab Development

Our students also showed a comprehensive set of skills and developed a cloud-based CAD system to support the teaching of ME303, and worked with me to design, build, and establish two teaching labs from an empty room to a space of self-expression and joyful learning. These tools are designed, used, and learned by our students.

ME336 Robotics & AI Guest Lecture Series

While teaching the lectures and leading the projects, I also invited a wide range of guest speakers to join our classroom, sharing their experiences when they were students of different majors and when they were developing industrial solutions for real-world problems, building a connection between theoretical development with industrial applications from a wide range of perspectives for students taking the class and the university.

Translating Course Projects into Publications

Students taking the ME336 and ME303 classes also showed continuous interest in further developing their course projects. They also had some new ideas while taking the class. Still, they were limited by the course structure and were interested in further exploring these ideas—many later published papers at leading conferences and journals in robotics and AI.

Student Competitions and Awards

Many of our students who took the course also participated in various competitions and won awards mainly in three ways. One is through further development of their course projects for competitions, the second is through developing teaching tools for classroom usage and then joining competitions, and the third is to join the SUSTech Robotics Club, where I serve as the advisor organizing competitions for students to participate. The example shown on the right is the waste sorting machine a group of students developed after taking ME303 in the Autumn, ME336 in the Spring, and ME491 in the Summer. Also, many students who participated in the DeepClaw project later joined the SUSTech Capstone Project for their final thesis and won the 1st, 2nd, and 3rd prizes in different years. Students who entered the robotics club also joined competitions such as RoboCon and others and won awards.

Student Achievement in Further Career Development:

Please refer to the list of students for more. Since joining SUSTech, 42 students graduated, including 37 Undergraduate and 5 Master Students, serving as their thesis supervisors, among which 20 for Jobs, 5 for Ph.D. study, and 17 for Master study. There are another 18 students currently under my supervision for thesis, including 7 Ph.D., 7 Master’s, and 4 Undergraduate Students. My former Ph.D. student graduated from Monash is currently a faculty member at the Huazhong Agriculture University.

Towards a Teaching Pedagogy of Design & Learning

I gradually developed a teaching pedagogy of my own after two stages of career development, which revolves around an interdisciplinary interaction between design and learning in teaching method, providing a link across fundamental knowledge, emerging technology, and industrial applications in course content, and creating a moment of joy to grow interests in the subjects the students are learning.

My First Experience with Teaching

During my post-doctoral period at the Singapore University of Technology and Design (SUTD), although I did not teach any classes, I got the opportunity to interact with many of SUTD’s students, where the feedback and discussion on their learning experience was always a topic that could last hours, providing me with a valuable perspective on the interaction between the instructor and the students. Also, I came across the concept of Project-based Learning (PBL) as a teaching method. During my post-doctoral period at the Massachusetts Institute of Technology (MIT), I joined the class of Product Design and Development lectured by Prof. Steven Eppinger, where a mixture of engineering students from MIT Mechanical, design students from Rhode Island School of Design, and business students from MIT Sloan, all at graduate levels, collaborate towards a themed design project within a tight schedule for creativity and development.

My Teaching Experience before Joining SUSTech

After becoming a Principal Investigator at Monash, I was assigned to teach ENG5002/6002 Engineering Entrepreneurship for master students, with only a few exceptions to undergraduates. The class size was about 70 when transferred to me, and then I developed the class to attract over 300 seats within two years. By transforming the class to an online format, I adopted a flipped classroom method to teach the course, invited external experts on entrepreneurship with various backgrounds as guest speakers, and introduced a PBL component to form 50+ student teams a semester for ideation, discussion, presentation, and prototyping. I modeled the final poster presentation session as a crowdfunding event where all students were given virtual credits to bid on their favorite project for virtual investment. It was an intense teaching and learning experience for me as an Early Career Researcher to be actively involved in the classroom and interact with students from a global perspective.

Courses I Taught at SUSTech

At SUSTech, I am the lead instructor of three courses, including ME303 Introduction to Mechanical Design (updated the course code to ME311 Mechanical Design) since the Autumn Semester of 2018, ME336 Collaborative Robot Learning since the Spring Semester of 2019, and ME491 Engineering Practice for the Summer Semester since 2020. Only ME303 and ME336 involve formal lecture hours, and the ME491 is a student-driven internship practice with no teaching duties. Also, collaboration with MIT was an active component of my teaching activities.

My teaching pedagogy that combines design and learning gradually emerged as I started teaching engineering subjects such as Mechanical Design and Robot Learning. Teaching them is challenging as these two subjects are positioned on two distinct ends for the Department of Mechanical and Energy Engineering students. The Mechanical Design course assigned to me is well-developed with classical textbooks, such as Shigley’s Mechanical Engineering Design, running up to the 10th edition already with mature standards and practices in the industry and academia. On the other hand, the Robot Learning course that I proposed is a new course that was not yet taught at all in China and an emerging subject in academic research with growing applications in the industry back in 2019. For the Mechanical Design class, I would give the students examples of how to formulate a data-driven understanding of the design of mechanical components and systems. For the Robot Learning class, I emphasize the importance of design thinking to help the students make sense of the machine learning techniques while dealing with collaborative robots with hands-on practices. For both courses, I practiced by inviting previous students taking my class to actively participate in iterating the lab projects for the next semester, making all course materials online, forming student teams for Project-based Learning, inviting guest speakers from the industry to join the classroom, adding presentation to the final class, and creating a joyful experience by making our students the stars of the course in the final project presentation. Such a teaching pedagogy takes a significant amount of my time during the semester but also provides me with closer access to the minds of SUSTech students, who are talented and curious but also mindful and cautious about interacting in classroom activities.

Teaching ME303 on Design and ME336 on Learning

Over the years, student enrollment in my class has shown a nearly five-fold increase in all courses I served as the lead instructor. During this time, the yearly registration of all students at SUSTech is 1,030 in 2018[1] and 1,330 in 2023[2]. There are multiple reasons for such positive change. Over the years, all courses underwent extensive changes and improvements with more optimized content offerings, schedule arrangement, and alignment with the student’s learning expectations. The department started a new major in Robotic Engineering in 2019 for undergraduate students on top of the existing major in Mechanical Engineering, attracting more students to study in the Department of Mechanical and Energy Engineering (MEE).

For ME303, a required unit for students within the department, I introduced significant changes to the course by adopting teaching practices learned from MIT in 2020, which shows a substantial increase in student enrollment. With the students’ growing interest in Robotic Engineering, enrollment in ME336, an elective course only intended for students majoring in Robotic Engineering, also showed a five-fold increase since its initial offering. Since 2020, the university has undergone many changes, with new departments offering the same major as those in MEE, resulting in a general spread of students enrolled at the MEE. It should be noted that the data shown below involves some of the same students taking different courses. Still, the general trend is positive to demonstrate the students’ interest and acknowledgment of my teaching activities at SUSTech since 2018.

The ME303 was offered only to undergraduate students since its very beginning. However, ME336 was initially intended for graduate and undergraduate students in the Spring of 2019. I updated the course to undergraduates only in the Spring of 2020 to avoid the mixing expectations from the graduate and undergraduate students.

ME303 Introduction to Mechanical Design

ME303 is a required course aimed at 3rd and 4th-year undergraduate students who majored in Mechanical Engineering and Robotic Engineering and optional for students who majored in Mechanics and Aerospace Engineering. This course is foundational for Mechanical Engineering students to train their mechanical design ability, teaching the basic knowledge, theory, and methods of design and skills to cultivate practical ability in engineering applications. It was a tremendous task while teaching this course, taking a significant amount of time during every Autumn Semester due to the foundational importance of this course as a pre-requisite of many other classes for all Mechanical and Robotic Engineering students. It also takes a significant amount in the Graduate Entrance Exam for undergraduates aiming to join SUSTech for graduate studies, which involves exam preparation at the national level. It is also practically essential to connect theoretical methods, engineering experience, national standards, and hands-on skills, all built upon SUSTech’s relatively limited development and accumulation in design-related subjects in teaching and research.

I took over this course in the Autumn Semester of 2018, right after I joined SUSTech in June. The first task assigned was to transform this course into an entire English course, following the university’s policy to promote a global environment for research and learning. Following the inauguration of the collaboration through the SUSTech-MIT Joint Centers for Mechanical Engineering Research and Education, an essential effort was made to translate teaching experience from MIT Mechanical to SUSTech, focusing on design-related courses. After reviewing the previous course outline of ME303 and the existing teaching lab at SUSTech, I introduced three significant changes to this course, with reflections summarized below that helped me better optimize the teaching and learning experience of this course.

Changing the textbook to Shigley’s Mechanical Engineering Design (10th edition), the same textbook used by the equivalent course at MIT, Elements of Mechanical Design, transformed the course into an entire English course. Preparing everything from ground zero is a massive task, teaching it with a new textbook in full English. Also, the conversion of metric systems and national standards became a problem I must resolve while teaching this course. In addition, due to the lack of mechanism-related subjects offered by the department, this course needs to cover both the engineering design of mechanical components and theoretical foundations on mechanism theory, adding too much burden to the comprehensiveness of course content. By then, the students needed to be fully prepared for an entire English course, making it difficult to absorb the lecture content effectively, especially when the English-Chinese translation was involved. Experience and feedback from this process are gradually optimized in the following years of this course to improve the learning experience.
Upgrading the design lab for teaching at SUSTech with modern equipment and safety measures. This critical component is necessary before teaching this course, as hands-on skills with modern equipment play a central role in helping the students learn the course. I applied for internal funding to fully renovate the previous lab space and equipment with modern tools, benches, machines, parts, and computers. I added protective safety components such as goggles, gloves, etc., to upgrade the lab, which is also a challenging task that is tight on schedule to design the lab, negotiate with contractors, purchase the equipment, and execute the construction within a very short time. The changes made throughout this process were later found essential for the department, making this design lab multi-functional, supporting a wide range of practical courses not limited to ME303 itself.
Introducing a project-based learning component in the course offering. Since this is the first time to add such a component, I mostly soft launch this change with an effortless project adapted from MIT’s course project, which students can complete within 1~2 weeks as a team.

The course remains the same in the Autumn Semester of 2019, with minor changes made in various places based on feedback and lessons learned from the Autumn Semester of 2018. During this semester, with the upgraded design lab, the project-based learning component was improved with a more comprehensive theme to design a robotic joint with dual outputs, and the student number almost doubled with the added students from Robotic Engineering.

Moving on to the Autumn Semester of 2020, the collaboration with MIT was gradually on track, and the general idea was to translate MIT’s successful experience in 2.009 by Prof. David Wallace to SUSTech by opening a new course coded ME499 Product Design and Development, with me as the lead instructor. The initial plan was to soft launch the course in the Summer of 2020 by inviting Prof. Wallace and his teaching team to SUSTech to coordinate a pilot run of this course by sharing the teaching method and course content so we could later adapt it for SUSTech. However, after reviewing the course content, teaching format, and teaching resources required for this course, I found this an even more challenging task, which requires significantly more resources than that was available at SUSTech, including more comprehensive preparation with a much larger lab space for project-based learning, an entire teaching support team covering various aspects of the hands-on process, and a theme-based project that is suitable to be carried out based on the curriculum landscape at SUSTech. The smooth delivery of this course also involves a dedicated software package to enable a practical feedback experience that is critically important for the communications between the teaching team and student teams.

To translate education excellence in emerging engineering from MIT’s 2.009 Product Engineering Processes, we experiment with the concept of large student teams with well-rounded teaching support for an interactive learning experience. While considering the differences between SUSTech and MIT in lab resources, teaching experiences, and student base, I proposed a year-long theme project series across three semesters in mechanical engineering, intelligent robotics, and engineering practice. I keep the lecture part unchanged to reflect the required fundamental learning but innovate in the lab projects through collaboration.

ME303 Introduction to Mechanical Design (Autumn Semester): a required core course toward a bachelor’s degree in mechanical engineering at SUSTech, focusing on the fundamental knowledge in engineering design and specifications of various mechanical elements used for modern machine design.
ME336 Collaborative Robot Learning (Spring Semester): an elective core course toward a bachelor’s degree in Robotic Engineering at SUSTech, focusing on integrating vision-based machine learning methods with collaborative robots for advanced manipulation problems in research and industry.
ME491 Engineering Practice / ME499 Engineer Product Development (Summer Semester): a required core course toward bachelor’s degrees in mechanical/Robotic Engineering at SUSTech, usually carried out by students taking internship assignments with research labs or related companies for hands-on projects in engineering.

Specifically, we changed the lab projects of the above courses through a theme project on “wasteless” to be carried out in three semesters by three different groups of students.

Machine Design in Autumn: a template production line of waste sorting is provided where three machines are to be designed by a team of ~15 students, including a splitting machine, a speedy picker, and a dex picker.
Robot Learning in Spring: now that the machines on the production line are ready, we focus on developing programs using vision-based machine learning algorithms to empower the production line with picking skills for waste sorting at high speed, accuracy, and intelligence.
Product Development in Summer: machine intelligence requires further integration with the engineering scenario to provide practical value through technology. We will visit the industrial need for waste sorting to integrate our system into an actual production line.

While preparing for this course, I worked with our students, who previously took this class, and developed several teaching tools to support this course, including a browser-based CAD system for rapid re-assembly of aluminum extrusion systems, a Feishu app to help online instant messaging with the students on teaching and learning feedback, a series of reconfigurable workstations for different courses, refurbished two lab spaces with equipment, tools, and parts for teaching activities, and a high-speed waste-sorting machine based on the student’s project design.

Although we tried our best to cope with COVID-19 during the Autumn Semester of 2020, continuous communication with MIT became significantly challenging to implement this teaching method at SUSTech. We had to abort the original plan of opening the new course of ME499 on Product Design and Development and stick with the existing infrastructure of ME303 with a lack of continuous support for the teaching staff and lab spaces. For the Autumn Semesters of 2021 and 2022, I gradually reversed the course to a more straightforward format with less concentration on students’ project involvement and increased the emphasis on an exam-based evaluation to suit the students’ growing need for the Graduate Entrance Exam. I updated the textbook to the Chinese version edited by the Huazhong University of Science and Technology, which holds a leading role in the teaching and research of Mechanical Engineering among universities in China. I updated the course delivery with dual-language support in 2021 and full Chinese in 2022. Starting in the Autumn Semester of 2023, all slides, textbooks, and teaching resources of ME303 will be converted to Chinese with a new course code ME311 on Mechanical Design without a lab or project component, focusing solely on practicing the design of mechanical parts.

ME336 Collaborative Robot Learning

This elective course is aimed at the 3rd and 4th year undergraduate students majoring in Robotic Engineering. The focus is on the problems of how a robot can learn to perceive the physical world well enough to act in it and make reliable plans. There are several challenges to teaching this course. This new subject is emerging in research with no formal textbook available for teaching. Stanford, UC Berkley, and Max Planck Institute are the first few universities teaching this course. From the robotics perspective, intersection with machine learning provides a data-driven method for computational intelligence in decision-making instead of model-based solutions. The development of machine learning research also seeks effective changes enabled by physical interactions with the external world for delivering the promised impact of artificial intelligence. The growing maturity in technology readiness levels of collaborative robots played a critical role in providing a reduced integration cost, safe interaction, and intuitive user interfaces for applied research with robot learning. Besides academic research, I surveyed the industrial demand for students with robot learning skills via various channels. Leading companies like Amazon showed tremendous interest by hosting the Amazon Picking Challenge contest at ICRA. Upon joining SUSTech, I was assigned as the director for a joint lab on logistic robotics with a startup showing similar interest. I also joined Team Nanyang, led by Prof. Chen I-Ming, one of my doctoral supervisors, to participate in a logistic robotic hosted by Jindong, where I experienced the industrial need for talents in this direction. Since this is also a research direction that I chose even back when I was a faculty member with Monash University, this could be an excellent opportunity to develop this new course on Collaborative Robot Learning at SUSTech. Besides the textbooks, other challenges, such as the need for more robotic hardware and severe limitations in teaching space for this course, also make it difficult to open this course at SUSTech. I conducted course development and made iterative changes over the years to design this course with a five-fold increase in student enrollment in the Spring of 2023, compared to the Spring of 2019 when this course was first offered at SUSTech.

When I started offering this course in the Spring of 2019, I intentionally limited student enrollment to 9 seats (a minimum requirement at the time for opening a course) due to the early development of this course, the limited robot workstations available at my lab (which were meant for research but had to be repositioned for teaching), and the lack of dedicated teaching space for students to safely operate these robots. Before the semester started, I built three sets of robot stations. I worked with my students to expand the original DeepClaw software package I developed at Monash for learning robotic picking with compatible interfaces for each robot workstation with different hardware. Considering the limited courses offered to SUSTech students in 2019 in the MEE department on subjects such as robotics, machine vision, and machine learning, I designed ME336 to cover a wide range of topics, including robotic collaboration, kinematics, Robot Operating Systems (ROS), robotic vision, calibration, RGB-D sensing, object recognition, artificial intelligence (AI) and deep learning (DL). Specific projects will be carried out throughout this course regarding the simulation of robot picking using fundamental kinematics and robot vision, an AI robot to play a tic-tac-toe game, and a DL robot to play an arcade claw game. As an elective course for a robotic engineering major, this course lays the foundation for students to use widely adopted Robot Operating Systems (ROS) to perform advanced robot control, including basic mathematical formulation, hardware usage, and intelligence integration. The following learning outcomes are expected for students taking this course:

Given functional and environmental requirements, utilize concept generation methods within a team setting to achieve a consensus for a robot design concept.
Design and develop functional robot programs from function, hardware, algorithm, and physical environment perspectives.
Apply the basics of disciplines, including mechanical engineering, electrical engineering, applied mathematics, and computer science, to understand the use of robots in action.
Communicate engineering decisions, justification for those decisions, designs, programming, and test results in multi-media presentation and report writing.

I intentionally designed the 1st version of the ME336 with an intense schedule. Besides formal lecture hours in theory and hands-on practice in my lab with the robot stations, I also invited guest speakers from the industry and academia to join our classroom and share their understanding and experience while developing industrial solutions using robot learning. The students received the course well with the upgraded version of the DeepClaw system in both hardware and software, hosted on GitHub for open-sourced access. It was also quite a learning experience for me while preparing and teaching this course and interacting with our students. I made three exciting observations by the end of this semester, asking the students for feedback on how to improve this course. One is the active participation from the undergraduate students rather than the master students, which inspired me to change this course to undergraduate only, who are keen on learning emerging topics in research and applications, whereas most master students seek only a pass. The second observation is the undergraduates’ keen interest in working on their projects during the upcoming summer, during which they would like to expand their projects with new ideas they came up with during the semester. The third observation is the undergraduate students’ diverse feedback on how this course could be improved in hardware, software, content, and project arrangement. As a result, I initiated a tradition with ME336 to invite the students after finishing this course and supervise further development of their projects into conference or journal publications or join me in the iterative development of the course’s lab project for the next batch of students. By the end of the summer, undergraduates of the class, including Yang Zeyi, Ge Sheng, Wang Haokun, and Liu Haiyuan, led or participated in the authorship of two journal articles and three conference papers through further development of what they have learned in the class for various research projects, which are the first publications at the BionicDL lab at SUSTech.

For the Spring Semester of 2020, the COVID-19 situation disrupted my original plan of ME336 with hands-on practice on the DeepClaw workstation, forcing us to switch to simulation-based projects for the students taking this course. For consistency with the previous course project, I invited Ge Sheng, who took ME336 in 2019 and stayed in my lab for research projects, to re-program the original course projects using V-Rep with changes that he felt were necessary from the student’s perspective while learning this course.

After this semester, I started to formally invite students to join me in the lab project development for ME336 after taking the course. This change started to pay off in the Spring of 2021, with student enrollment increasing nearly 4-fold to 35, partially due to the temporary lab space for teaching inherited from ME303 in the Autumn of 2020. But we were still limited to three robot workstations back then. However, the students showed a strong interest in learning new subjects in robot learning while forming teams to improve the machines they designed in ME303 into a working solution. They developed four robotic systems with different involvement of learning-based autonomy in decision-making and physical interactions for waste sorting, including one for robotic trash tossing into bins based on object classification using machine vision, one for waste classification based on tactile sensing, one for imitation learning of waste picking for robotic arms, and one for ambient intelligence for safety assurance in human-robot interaction.

For the Spring Semester of ME336 in 2022, I made a series of course delivery changes for multiple reasons. One is the continuous uncertainties caused by COVID-19, making it a mystery whether we can teach in the classroom or through video conferencing. The other is the transfer of the teaching lab to the department, meaning that I must make significant changes to the lab session to ensure the same learning experience and teaching outcomes. After communication with our students, we decided to use the summer of 2021 to develop a portable version of the DeepClaw system at a much-reduced cost for mass adoption in the classroom without direct involvement with the limited three sets of robot workstation at my lab, which is also desperately needed by the graduate students at my lab for daily research activities. Three students, Dong Yujian, Xiao Yang, and Wei Jinqi, who took ME336 in the spring semester of 2021, were interested in this project and joined the development of the portable DeepClaw, which is based on a prototype design reported in a draft paper written by Wang Haokun, a previous student of ME336 in the spring of 2019.

By the end of the summer, they designed three versions of the portable DeepClaw, including a fanny pack version with the lowest cost, a desktop version with an inverted projector, and a force-feedback version for hand-held operation. Based on a balanced consideration of the complexity of development, fabrication cost, space limitation, and tight schedule of preparation before the spring of 2022, we decided to move on with the fanny pack version for further development. We fabricated a total of 40 sets of the portable DeepClaw, designed a pair of kitchen tongs with two soft fingers attached to the end, connected to an external camera capturing the spatial deformation of the soft fingers and tong movement initiated by the student’s hands, and provided a browser-based interface for data collection without coding in the Linux environment. With the portable DeepClaw, we could expand the enrollment for more students without needing a dedicated teaching lab, as the students could directly operate with the portable DeepClaw to learn and practice some of the critical elements involved during a robot learning project. Due to the positive feedback on the portable DeepClaw, we kept using them for the spring semester of 2023 with updated tutorial sessions and materials posted on the course website for open-sourced access. Recently, we submitted a conference paper to ICARM2024 to report our development of the portable DeepClaw system in research and teaching.

Engineering Practice with ME491 and Others

This is a practice-based course for students to participate in self-selected projects with research labs or industrial collaborators and prepare a report for my assessment and approval by the end. My function is to review student applications, assess their reports, and provide accreditation approval by the end. Various courses are coded in the ME49X series, and ME491 is one targeted explicitly for Robotic Engineering Students. I was also involved as a supervisor for students in my lab taking other ME49X courses.

Machine Intelligence in Design and Learning

Looking into the future of my teaching activities at SUSTech, the focus is on Textbook Writing, Teaching Tool Development, and continuous support for student success.

For ME303, the university still encourages the need for dual language education with some of our students from foreign countries. While there are many English textbooks about Mechanical Design, there is no English textbook on this subject based on Chinese standards, especially when the national standards and metric system are involved. Pursuing this direction for a globalized educational effort could be helpful, but it would require more time and preparation before starting writing.

For ME336, while Robot Learning is rapidly evolving, it would be beneficial to write a textbook on this subject that suits my teaching activities at SUSTech. This has been on my plan since I joined SUSTech. Still, with the development of DeepClaw, I’m more interested in combining both into a shareable and reproducible teaching tool with online learning resources as a more suitable direction to pursue, which I plan to further develop within the next five years.

Finally, the “success” of my teaching is ultimately reflected by the students’ success through an in-depth understanding of their knowledge and skills, their personal career development, and their original and independent realization of their aspirations for the future.

[1] https://zs.sustech.edu.cn/page/563

[2] https://newshub.sustech.edu.cn/html/202307/44181.html