Pito Salas

Seven years ago (2016), we began integrating Robotics into our Com- puter Science curriculum. This paper explores the mission, initial goals and ob- jectives, specific choices we made along the way, and why and outcomes. Of course, we were not the first to do so. Our contribution in this paper is to describe a seven-year experience in the hope that others going down this road will benefit, perhaps avoiding some missteps and dead-ends. We offer our answers to many questions that anyone undertaking bootstrapping a new robotics program may have to deal with. At the end of the paper, we discuss a set of lessons learned, including striking the right balance between depth and breadth in syllabus design and material organization, the significance of utilizing physical robots and crite- ria for selecting a suitable robotics platform, insights into the scope and design of a robotics lab, the necessity of standardizing hardware and software configu- rations, along with implementation methods, and strategies for preparing students for the steep learning curve.

In this Work-In-Progress paper, we describe our vision for a remote/virtual robotics lab and our progress towards realizing it. Robotics has been taught at our institution for four years with enrollment growing each year. Our program is part of the undergraduate Bachelor of Science in Computer Science and has an equal focus on robotics theory and practice, with the “Robot Operating System” (ROS) running on students' own computers, and a small robotics learning lab with a variety of ROS based robots. Our pandemic experience during 2020 led us to invest in a cloud-based shared “cloud” environment running on a Kubernetes cluster. We learned that there are real advantages to this in terms of scale and geographical reach. The success of this with our students, our Covid experience, preparation for teaching in our lab, and planning towards return to campus led directly to this work-in-progress paper, where we define, in one abstraction, a remote/virtual robotics learning lab, unifying local and remote access to both simulated and real robots. This work aims to extend our existing shared virtual cloud-based robotics learning environment and combine it with a remote-controlled robotics lab.

The Robotics program at our institution has had growing enrollment over the past four years. This undergraduate program has an equal focus on theory and practice, with students doing hands-on work in our Robotics Lab. Last year we, like many others, faced the challenge of continuing teaching Robotics to newly remote learners. This paper reports on our response to this challenge, which was to embark on a project to create a robust, scalable, student-friendly and cost-effective virtual robotics lab. We describe a series of assumptions that we had and how each one proved out, conclude with our recommended architecture which (perhaps surprisingly) does not use any of the many existing cloud services. Instead it relies on a cluster of inexpensive “game” servers.

We find that this approach supports all our requirements, which were:
a) to provide a uniform experience for students on any operating system;
b) to mimic as much as possible the student experience in the physical lab.
c) to be able to supervise and assist students while they were using the lab; and
d) to support the growing number of students and
e) to be cost-effective.

Our recommended system meets and exceeds all of these requirements in the following way: Each student is provided with a private virtual environment which has the required operating system, packages, scripts pre-installed. These environments, managed by Kubernetes, can easily be created and recreated if needed. The student can access the virtual desktop, simulation environment and programming environment with a standard web browser on any platform. Provisioning new students, new software, homework assignments is semi-automated and scales well. In one semester of intensive use we have had high reliability, low cost and very high student satisfaction.

This paper describes three years of experience with an intensive three course summer semester on web and mobile entrepreneurship for second year CS students and beyond. The program is similar in structure to a high school summer camp or to a summer accelerator/incubator program except that it has a much higher level of academic content and provides the credit equivalent to three Computer Science electives and a full semester of residency. The program has been effective at teaching students production programming and entrepreneurship and has stimulated entrepreneurial activity during the academic year. It is taken by about half of the CS majors in the department which is surprising since it requires them to spend their summer months in a very intense academic and entrepreneurial experience which is quite different from the usual summer experience of their peers.