Visible to the public File preview

Towards a Unified CPS Education: Lessons Learned from a Cross-Disciplinary Robotics Engineering Program
Taskin Padir and Michael A. Gennert
Robotics Engineering Program Worcester Polytechnic Institute {tpadir, michaelg}@wpi.edu
Abstract—In 2007, Worcester Polytechnic Institute (WPI) introduced an undergraduate degree program in robotics engineering, first in the United States, with the goal of educating a new cadre of engineering students with multidisciplinary skills to meet the workforce needs of the growing robotics industry. A four-course core curriculum called unified robotics is at the center of the program and provides an effective way of teaching foundations of robotics engineering from traditional disciplines of computer science, electrical engineering and mechanical engineering. In its fifth year, the program has been highly successful in meeting its educational outcomes in terms of quantity and quality of enrolled students, ABET EAC accreditation, placement in engineering workforce and graduate school, and course and project evaluations. In this paper, we present our approach to designing a new cross-disciplinary engineering degree program, share lessons learned from several revisions of the program and discuss the synergy and common traits between cyber-physical systems and robotics education.

I. I NTRODUCTION Over the centuries, the engineering profession has evolved to meet the growing needs of society. As a result, we witnessed the demand and creation of new engineering disciplines. Biomedical, environmental, aeronautical and computer engineering are only a few examples of engineering branches that emerged to educate the next generations of engineers to find solutions to new challenges requiring a diverse set of knowledge and expertise. A recent report by UNESCO [1] provides a new perspective on the importance of engineering in development. According to the report, a new wave of engineering innovation is currently taking place, incorporating concepts such as sustainability, renewable energy, resource productivity, biomimicry and whole system design. Cyber-physical systems (CPS) are whole systems whose reliable, safe and efficient operation rely upon a tight integration of cyber (computation) and physical (sensing, actuation) domains [2]. Science, engineering, and technology of CPS is on the verge of rapid development and hence, there is a tremendous need to educate the next generation of scientists and engineers who have cross-disciplinary knowledge and training to engineer the CPS of tomorrow. Motivated by the growing needs of robotics industry, the unstoppable enthusiasm among the current generation of high school students, and the resulting growth in number of students in STEM fields, Worcester Polytechnic Institute (WPI) introduced an undergraduate Bachelor of Science (B.S.) degree program in Robotics Engineering (RBE) in 2007 [3].

The robotics engineering faculty adopted as a vision the creation of an exemplary, nationally recognized, multidisciplinary center for education, research, and innovation in robotics. The primary goal of the program is to educate engineers for the 21st century, the enterprising engineers envisioned by Tryggvason and Apelian, who “knows everything, can do anything, collaborates, and innovates.” There is a great deal of synergy between CPS and robotics, the integration of sensing, computing and actuation in the physical world. Indeed many CPS incorporates robotic systems for interaction with the physical environment. Furthermore, one can argue that an intelligent robot itself is a CPS [5]. Both robotics and CPS are multidisciplinary fields; creating a robot or CPS requires a whole system design approach; both promote and require teamwork, technical competency, innovation and lifelong learning. We posit that WPI’s 5year experience with its ABET–accredited robotics engineering program will provide a model for CPS educators in designing and implementing a cross-disciplinary program. The paper is organized as follows. Section II provides an overview of the unified robotics program. The details and practical content of an example course is presented in Section III. Section IV briefly discusses a sample robotics capstone project which follows a CPS design methodology. Finally, Section V provides a discussion of the lessons learned and our findings on student learning as well as a vision for a unified CPS curriculum. II. U NIFIED ROBOTICS C URRICULUM The Robotics Engineering program integrates foundational concepts from computer science, electrical and computer engineering, and mechanical engineering to introduce students to the multidisciplinary theory and practice of robotics. The program aims to provide students with both the disciplinary fundamentals and interdisciplinary outlook needed for success in this dynamic and growing field. The educational program objectives and outcomes are as follows. Educational Program Objectives: The robotics engineering program strives to educate men and women to: 1) Have a basic understanding of the fundamentals of Computer Science, Electrical and Computer Engineering, Mechanical Engineering, and Systems Engineering. 2) Apply these abstract concepts and practical skills to design and construct robots and robotic systems for diverse applications.

Fig. 1. The WPI Robotics Engineering program is structured around a core consisting of Introduction to Robotics, Unified Robotics I-IV, and the Capstone Project [6].

3) Have the imagination to see how robotics can be used to improve society and the entrepreneurial background and spirit to make their ideas become reality. 4) Demonstrate the ethical behavior and standards expected of responsible professionals functioning in a diverse society. Educational Outcomes: Graduating students will have: a. an ability to apply broad knowledge of mathematics, science, and engineering, b. an ability to design and conduct experiments, as well as to analyze and interpret data, c. an ability to design a robotic system, component, or process to meet desired needs within realistic constraints such as economic, environmental, social, political, ethical, health and safety, manufacturability, and sustainability, d. an ability to function on multi-disciplinary teams, e. an ability to identify, formulate, and solve engineering problems, f. an understanding of professional and ethical responsibility, g. an ability to communicate effectively, h. the broad education necessary to understand the impact of engineering solutions in a global, economic, environmental, and societal context, i. a recognition of the need for, and an ability to engage in life-long learning, j. a knowledge of contemporary issues, and k. an ability to use the techniques, skills, and modern engineering tools necessary for engineering practice. The educational outcomes are aligned with ABET EAC criteria, as a result, the program was granted accreditation under General Engineering criteria in 2011. In order to meet the educational outcomes, a series of undergraduate courses were created consisting of an introductory engineering class with a robotics focus at the freshman level and a four-course Unified Robotics sequence at the sophomore and junior levels. Figure 1 provides a visualization of the RBE curriculum. The Unified Robotics I-IV course sequence forms the core

of the robotics engineering program at WPI. The sophomore level courses, RBE 2001 Unified Robotics I: Actuation and RBE 2002 Unified Robotics II: Sensing, introduce students to the foundational concepts of robotics engineering such as mechanisms, position and velocity analysis, stress and strain, pneumatics, circuits, operational amplifiers, electric motors and motor drive circuits, sensors, signal conditioning and embedded system programming using the C programming language. The goal is to introduce students to the analysis of electrical and mechanical systems as well as the principles of software engineering. The junior level courses, RBE 3001 Unified Robotics III: Manipulation and RBE 3002 Unified Robotics IV: Navigation, build on this foundation to ensure that students understand the analysis of selected components and learn system-level design and development of a robotic system including embedded design. These junior-level courses provide a much deeper theoretical coverage of robotics, including: coordinate systems and frame transformations, manipulator kinematics and dynamics, modeling and control, sensors, signals, reasoning with uncertainty, navigation, mapping and path planning. The design of content for the unified robotics curriculum has been motivated by and attempts to address the following questions: (1) What comprises a meaningful laboratory experience in an undergraduate robotics course? (2) What is the appropriate level of robotics education for undergraduates? (3) How do we ensure that students can reach a level of robotics theory and practice to accomplish a reasonable yet satisfying course project? (4) How do we maintain student interest and learning active as the courses progress? (5) What is the required content for these courses that will lead into a comprehensive robotics capstone experience? The approach adopted by the faculty within this framework can be described by the following common features: 1) Each course culminates with a comprehensive course project that directly builds on the hardware and software modules and experience from the previous lab experiments. 2) Structured and well-defined lab exercises provide students with the necessary in-depth knowledge and practice to prepare them for the course project. 3) Lectures and homework assignments are designed to support the individual lab experiments and eventually the course projects. 4) A systems level approach to design has been adopted in each course. In this approach, designing a robot for a specific functionality requires an understanding of the user requirements, developing a set of design specifications, an integrated design effort to meet the specifications, and a design validation process. All courses are offered in 7-week terms with 4 hours of lecture and 2 hours of in-laboratory instruction per week. WPI students are expected to spend approximately 17 hours per week per course within the 7-week term structure. The typical student course load is 3 courses per term. Furthermore, in keeping with the long history of the WPI Plan [7], all courses emphasize project-based learning in small teams, hands-on and open-ended assignments, and students’ commitment to learning outside the classroom. Robotics courses provide an excellent opportunity for implementing this project-based instructional approach by incorporating open-ended projects with

detailed timelines and milestones. III. E XAMPLE 1– U NIFIED ROBOTICS IV: NAVIGATION The focus in Unified Robotics IV: Navigation is on integrating the information students acquired in their prior courses into a complex robotic system. The emphasis shifts to higherlevel programming, intelligence and algorithms implemented on a mobile robot platform. Upon successful completion of this course, students are able to: 1) Compute mobile robot kinematics. 2) Develop a model for mobile robot platform dynamics. 3) Develop a distributed architecture mobile robotic system. 4) Implement navigation algorithms based on sensor fusion and environment representation. 5) Write moderately involved programs in Java to control real-time tasks with a robotic system. 6) Construct, program, and test the operation of a robotic system to perform a specified task. Lab Assignments and Course Project We have developed four laboratory assignments and a multi-week course project in order to fulfill the course objectives outlined above. Each lab and the project builds upon the work completed in previous labs. Students work in teams of 2 or 3 on all lab and project assignments. The content of the labs and project are as follows: 1) Lab: Software Framework–Students are introduced to Unified Robotics IV software framework and begin to develop a distributed, object-oriented software system for the mobile robot platform which is used throughout the course. 2) Lab: Kinematics and Odometry–Students develop a kinematics model for their mobile robot platform and implement methods for differential drive motion, simple trajectory generation (drive along a straight-line or a circular arc) and odometry calculation. 3) Lab: Path Planning–Students implement an occupancy grid based path planning algorithm and path traversal through waypoint navigation. The lab objective is to demonstrate the robot’s ability to plan and follow a path within a known world map. 4) Lab: Mapping–Students implement mapping algorithms based on occupancy grid and line map representations, as well as a visualization tool to display the map data. 5) Course Project: Autonomous Mapping Robot–Students program a mobile robot to autonomously navigate in an unknown environment while mapping its environment. The focus is on high-level mapping and navigation tasks to create a world model of a maze and then travel through it. The project emphasizes mobile robot motion control, obstacle avoidance and navigation planning at the high-level and complements the projects student complete in the previous RBE courses. All the knowledge and practical experience gained by the students in the Unified Robotics IV lab assignments is applicable to completing the course project. The project typically covers a duration of three weeks during which the progress of each project team is monitored by weekly design reviews. The openended course project allows students to think outside the box and tackle the problem from a whole system design point of view. Within this framework, students successfully implement complex autonomous robot navigation algorithms including

Fig. 2. Software architecture developed in a capstone robotics project for a semi-autonomous wheelchair used in a human-in-the-loop CPS demonstrating tight integration between the physical and computational domains.

probabilistic occupancy grids, extended Kalman filters, particle filters and path planning methods. IV. E XAMPLE 2– ROBOTICS C APSTONE Robotics capstone is equivalent to three courses and typically completed in one academic year. We present a case study to demonstrate the commonalities between robotics and CPS projects. In this project, the goal is to design and realize a sensor network and corresponding intelligence and autonomy modules to convert a commercially available powered wheelchair into a semi-autonomous mobility platform. A team of three RBE senior students designed, developed, and validated a sensor network including multimodal (including 1 infrared, 1 ultrasonic) range sensor modules, a communications module, a vision based environment mapping module and corresponding software firmware as well as high-level algorithms. Students had to pay specific attention to integration, safety, reliability, modularity and interoperability. The software architecture developed within the scope of this project is depicted in Figure 2 demonstrating the tight integration of physical and computational components. This project can be viewed as a mini-scale CPS project. V. D ISCUSSION The following observations can be made about the unified robotics curriculum in relation to CPS education: • Robotics in nature is a multidisciplinary engineering field similar to CPS, therefore it needs to be learned in a unified and integrated manner. Robotics, similar to CPS, is a complicated field to be taught at the undergraduate level. The course content needs to be carefully designed and supported to ensure that the students receive adequate breadth and depth. Student experience can be summarized by the following statement from a course evaluation: “We work hard, the courses are hard, but we learn a lot. So, it is worth it.”. 80% of the students reported that they spent more than 17 hours on their coursework. Moreover, they ranked the amount that they learned and the intellectual challenge presented by the course as 4.8 out of 5.0 which is an indicator for student













satisfaction. It is expected that a comprehensive CPS education will require a similar effort level. It should be noted that the curriculum includes other robotics courses and a capstone design project. This provides ample opportunities for instructors to emphasize the mechanical, electrical and software design concepts throughout the program. Unified robotics curriculum is spread over multiple years and self-contained to the extend that students are required to take foundational math, science and engineering courses. The freshmen level introduction course provides an effective venue to introduce students to the cross-disciplinary skills expected from them during their study as well as to keep them engaged in robotics. The program structure is flexible enough to accommodate students who do not decide “early” that they would like to pursue the program. Students can decide to take the robotics curriculum as late as the second semester of their sophomore year and can still complete the program on time. Students who are not in the program can still be exposed to robotics topics either by pursuing a minor or taking any of the RBE courses as their engineering electives provided they have the recommended background.

Fig. 3.

A proposed design approach for CPS education.

attractive to students. The goal is to keep students engaged in active learning of complex theories and tools for CPS. The students attain the learning objectives of the project by building their knowledge through well-structured lab exercises whose outcomes are tightly related to the project outcomes. Finally, lectures and homework assignments emphasize the foundational theories of CPS. This approach is scalable from undergraduate to graduate courses, and adaptable to various CPS application domains. VI. C ONCLUSION We presented our approach to designing a crossdisciplinary robotics engineering degree program and shared lessons learned. The synergy between robotics and CPS can potentially enable CPS educators to adopt the curriculum design methodology discussed in this paper. ACKNOWLEDGMENT The authors would like to thank their students for their feedback in creating and revising the unified robotics curriculum. This material is based upon work supported by the National Science Foundation under Grant No. 1135854.
[1]

A. Lessons Learned Several important lessons emerge from 5 years experience with Robotics Engineering at WPI. First, robotics engineering is a viable major, attracting students from a wide geographic area. Not only does it bring students in, but they graduate to successful positions. A robotics engineering program can be accredited by ABET, providing some additional assurance of its academic merit. A key factor in the success of the program is the collaboration of faculty and staff from different departments, reporting to different deans, and the support of the administration. Throughout the program’s development, there was a free and natural exchange of ideas, and no one of the supporting departments dominated the others. To the contrary, every effort was made to accommodate departmental differences, incorporating the best aspects of each. The curriculum and the courses generally proceed quite well, although they are challenging to teach, and care must be taken that each courses runs as smoothly as possible. However, there is a steady amount of tinkering that must be done in the curriculum, in the syllabi, with hardware, and in software, as experience is gained. There are more than 100 graduates of the program to date and 95% of all were known to be working or in graduate school. Approximately 1/3 of these graduates attend graduate school, 1/3 work in the robotics industry, and 1/3 work in engineering not specifically in robotics. B. Unified CPS Curriculum Motivated by the WPI’s unified robotics curriculum, a similar approach can be adopted to design cross-disciplinary CPS courses and curricula. Figure 3 presents the proposed design process for the educational experiences in CPS. Each learning activity (module, course, complete curricula) is motivated by a level-appropriate CPS project from an appropriate application domain (healthcare, energy, manufacturing, etc.)

[2] [3]

[4]

[5]

[6]

[7]

R EFERENCES UNESCO Report, “Engineering: Issues, Challenges and Opportunities for Development”, 2010. http://unesdoc.unesco.org/images/0018/001897/189753e.pdf, (Last accessed in February 2013) Baheti, R. and Gill, H. “Cyber-physical systems,” The Impact of Control Technology 161-166, 2011. Padir, T., Gennert, M.A., Fischer, G., Michalson, W.R. and Cobb, E.C., “Implementation of an Undergraduate Robotics Engineering Curriculum”, ASEE Computers in Education Journal, Special Issue on Novel Approaches to Robotics Education, vol. 1, no. 3, pp. 92-101, 2010. Tryggvason, G. and Apelian, D., “Re-Engineering Engineering Education for the Challenges of the 21st Century,” Commentary in JOM: The Member Journal of TMS, Oct. 2006. Lee, E.A and Seshia, S.A., “Introduction to Embedded Systems, A Cyber-Physical Systems Approach,” http://LeeSeshia.org, ISBN 978-0557-70857-4, 2011. Padir, T., Fischer, G., Chernova, S. and Gennert, M.A., “A Unified and Integrated Approach to Teaching a Two-Course Sequence in Robotics Engineering,” Journal of Robotics and Mechatronics, Vol. 23 No. 5, 2011. “The WPI Plan: 40 Years of Innovation and Counting,” http://www.wpi.edu/news/perspectives/108116.htm, (Last accessed in February 2013).