About Peter Chipman

I'm a lexicographer, an editor, and a lover of language and literature. Also a proud father of two, an occasional bell-ringer, a thirteenth-generation New England Yankee, a former owner of a one-room schoolhouse, and the current owner of a 220-year-old farmhouse.

Doctoral Students Aren’t Lone Wolves: An interview with Brian Charles Williams

By Peter Chipman, Digital Publication Specialist and OCW Educator Assistant

The Curiosity rover standing on the surface of Mars

The Curiosity Mars rover, a complex, collaboratively-built system based on cognitive robotics. (Image credit: NASA/JPL-Caltech/MSSS)

Robotics and artificial intelligence are fast-paced fields in which researchers constantly have to adapt to new technological developments. But in such fields, progress isn’t always achieved by competitive, individual effort; in many circumstances, cooperation and collaboration are more fruitful approaches. In the interview excerpt below, Brian Charles Williams, a professor at MIT’s Computer Science & Artificial Intelligence Laboratory, describes how he develops learning communities in the graduate-level course 16.412 Cognitive Robotics:

OCW: How is learning different in a course focused on an emerging field like cognitive robotics?

Brian Williams: Students are accustomed to reading chapters in textbooks—material that took decades for scientists to understand. But cognitive robotics is an active research area. It’s moving so quickly that every three years or so it reinvents itself. This course is focused on helping students close the gaps in the research. To be at the cutting edge of research, students need to read across papers and understand core ideas that are developed from a collection of publications. And then they need to be able to reduce that understanding to practice.

There’s also no better way to understand something than to teach it, implement it, and put it in a bigger context of some real-world application. That’s why we have a grand challenge at the center of the course experience.

OCW: Tell us more about the grand challenge.

Brian Williams: I like the idea of learning communities, of everybody trying to learn about a topic together. The grand challenge is a communal learning experience driven by a cutting-edge research question in cognitive robotics that allows us to focus on core reasoning algorithms. Students work in teams to present advanced lectures about different aspects of the topic.

OCW: Why teams?

Brian Williams: It’s important for students to work in teams because research is a collaborative endeavor. The notion that doctoral students are lone wolves is just not accurate. The more students can practice effective collaboration, the better.

It’s also the case that developing lectures is hard work. Just producing a first draft of a lecture can take 20 to 30 hours. And then you need to spend another 6 hours improving it. So, to develop a high-quality lecture, you really need two people working together.

Robot standing in a room.

Domo Robot, developed by Aron Edsinger and Jeff Weber, is able to adapt to novel situations. It is on display in the MIT Computer Science and Artificial Intelligence Lab. (Image by MIT OpenCourseWare.)

OCW: How do you assess student work completed collaboratively?

Brian Williams: That is an interesting problem, because when the whole class does a project collaboratively teams can become too large. When that happens, people begin to feel disenfranchised. What I do to combat that is to make clear from the beginning what elements or materials individuals are responsible for contributing to the project. I have students write down what they are contributing so that I can assess their work accurately.

Another piece of the assessment puzzle is providing good feedback. The place where feedback matters the most is during the dry run for the students’ advanced lectures. A week before the students give their lecture to the class, they do a dry run for the teaching team and receive feedback. The process takes about two and a half hours. We teach them how to capture students’ interest at the beginning of the lecture and how to clarify the main points they want students to learn. We also help them convey the synergies between the main points and encourage them to consider the role of examples in their presentations.

OCW: Are there other components of the grand challenge, in addition to the advanced lectures?

Brian Williams: Yes. As I mentioned, the field of cognitive robotics is moving really fast. What normally happens is that members of the research community will generate tutorials on emerging themes. These tutorials encapsulate core ideas that everybody should know. The problem is that there’s just so much we need to know—but not enough time to write all the tutorials. So some of the students in the class are assigned to write tutorials related to the topic of the grand challenge. And a few others will write corresponding Jupiter or Python notebook problem sets. Along with the lectures, students end up producing materials that are enormously helpful to researchers in the field. This is important because I want them to learn that as scientists, their role is to consolidate ideas and to teach the community.

Man sitting at desk. Bookshelves with books to his left.

Aeronautics and Astronautics Professor Brian Williams in his office on the MIT campus. (Image by MIT OpenCourseWare.)

OCW: It’s interesting that you have the goal of figuring out cognitive robotics as a field, but also how to teach it to others.

Brian Williams: And how to catalyze community. An engaged, collaborative community is absolutely key.


You can read more of Professor Williams’s thoughts about teaching 16.412 on the Instructor Insights page of this course.

Keep learning! The following courses and Instructor Insights may be of interest to you:

Another OCW Course Offered by Professor Williams

Image combining data taken by an autonomous vehicle with the views from its windows.16.410 Principles of Autonomy and Decision Making

This course surveys a variety of reasoning, optimization, and decision making methodologies for creating highly autonomous systems and decision support aids. The focus is on principles, algorithms, and their application, taken from the disciplines of artificial intelligence and operations research.

Reasoning paradigms include logic and deduction, heuristic and constraint-based search, model-based reasoning, planning and execution, and machine learning. Optimization paradigms include linear programming, integer programming, and dynamic programming. Decision-making paradigms include decision theoretic planning, and Markov decision processes.

More about Robotics and Artificial Intelligence

Wheeled robot carrying doll in its arm.2.12 Introduction to Robotics

This course provides an overview of robot mechanisms, dynamics, and intelligent controls. Topics include planar and spatial kinematics, and motion planning; mechanism design for manipulators and mobile robots, multi-rigid-body dynamics, 3D graphic simulation; control design, actuators, and sensors; wireless networking, task modeling, human-machine interface, and embedded software. Weekly laboratories provide experience with servo drives, real-time control, and embedded software. Students design and fabricate working robotic systems in a group-based term project.

Graphic of three figures in an evolutionary arc, starting with a figure standing upright on the left, ending with a person hunched over at a computer on the right.6.034 Artificial Intelligence

This course introduces students to the basic knowledge representation, problem solving, and learning methods of artificial intelligence. Upon completion of 6.034, students should be able to develop intelligent systems by assembling solutions to concrete computational problems; understand the role of knowledge representation, problem solving, and learning in intelligent-system engineering; and appreciate the role of problem solving, vision, and language in understanding human intelligence from a computational perspective.

More on Learning Communities

Illustration of a brain with colors indicating regions involved in social processes, plus three example faces used in social testing and a photo of a large group sitting in a circle.9.70 Social Psychology

In the rather idiosyncratic syllabus for this course, which goes into much more philosophical depth than such documents usually attain, Professor Stephan L. Chorover lays out the principles of the collaborative learning system that formed the basis of his approach to teaching.

A woman on the monitor of a video camera, facing the viewer20.219 Becoming the Next Bill Nye: Writing and Hosting the Educational Show

Elizabeth Choe and Jaime Goldstein discuss the importance of cultivating a sense of community in the classroom, and explain how situating themselves as facilitators-of-learning, rather than omniscient givers-of-knowledge, communicated to students their respect for them as learners.

The faces of the four members of the Beatles in a 4 by 4 grid.21M.299 The Beatles

The Beatles lived an insulated life in the 1960s. They couldn’t go out without being mobbed. As a result, the four of them spent much of their time together, listening to and playing music. In that process, they were constantly learning from each other. Lecturer Teresa Neff discusses the centrality of group learning in her Instructor Insights for this course.

Find insights like these on many other teaching approaches at our Educator Portal.

How Would You Like Your Grade? An Interview with George Verghese

By Peter Chipman, Digital Publication Specialist and OCW Educator Assistant

black lines, indicating elements of heart beats, on red and white graph paper.

Electrocardiogram data, an example of measured signals. (Image courtesy of kenfagerdotcom on flickr. License: CC BY-NC-SA.)

The syllabus for a typical MIT course spells out a familiar grading scheme that assigns fixed percentage weights to the different elements of the course: so many points for attendance and participation, so many for the quizzes or written assignments, and so many for the final exam or final project. Such a system is straightforward to implement and easy for students to understand, but there are times when both students and instructors want a little more flexibility. After all, not all students are the same, and they don’t all have the same needs.

In Spring 2018, Professors George Verghese, Alan V. Oppenheim, and Peter Hagelstein co-taught 6.011 Signals, Systems and Inference, an undergraduate course that covers a broad range of topics pertaining to communication, control, and signal processing. The material is complex, and the instructors support student learning in unique ways. We approached Professor Verghese for his insights on the course’s unusual grading system, and also on how the teaching team uses tutorials and an informal collaborative learning space called the Common Room to help MIT students succeed.

OCW: You offer three grading schemes in the course: regular, lower-friction, and project. This is so interesting! Tell us about your decision to offer students this kind of choice.

George Verghese: Ideally we’d like all students to attend all lectures and recitations, and for most students this is essential to their learning the material well and succeeding in the class. However, student lives can be complicated, their backgrounds and motivations are varied, and they optimize their trajectories through MIT in different ways. I know there is always a handful of students who can master the material with much less interaction, and I am fine with allowing them to do that, without getting in their way—hence the “lower-friction option,” in which the only components of the grade are their scores on the homework, the two quizzes during the term, and on the final exam. Students who opt for this have access to all the material in the class, but not to the tutorial sessions, as we don’t want them using the teaching assistants to help them make up for lecture and recitation material they may have missed. Unfortunately, there are always a couple of students who opt for the lower-friction version who really should not have, and their grade ends up suffering for it. But there are others—and these are the ones for whom this option is intended—who end up near, or even at, the top of the class. More power to them! My only regret is that we lose the benefit of whatever they may have contributed to class discussions if they had attended lectures and recitations.

Those students who do not elect the lower-friction option are, in effect, signing on to attending most lectures and recitations, and 15% of their course grade is allotted to attendance. I don’t actually take attendance directly, but every few lectures I will have them pair up in class to work out some problem related to the lecture material, then turn in their answer sheets at the end of lecture (with their neighbor’s name on their sheet, so they know I’m not looking to grade them on their answers!). Any student who misses a couple of these gets a note from me to urge better attendance. And recitation instructors have a good sense of who is attending and who isn’t, even if they don’t take attendance formally.

Lecture 22 image

An example of binary hypothesis testing from Lecture 22 (PDF).

There are also a few students each semester who feel they’d do better if they had a project to anchor their learning, and also to spread the course grade over (10% of the course grade is assigned to the project for students who choose this option, and the contributions of quizzes and the final exam are correspondingly reduced). Since there are typically only a few such students, I work quite closely with them over the semester, meeting at least every couple of weeks, to ensure the projects are related to course material and are moving along well. Some of these projects turn out to be good demonstrations for lectures in succeeding terms.

OCW: Please describe the tutorials offered to students and tell us about their role in the course.

George Verghese: Our tutorials are run by the teaching assistants on an optional, sign-up basis, limited to 5 students per session. Some students—perhaps a third of the class—attend them very regularly each week, others occasionally or not at all. The idea here is to actively engage the students, have them go the board to work things out, rather than having the teaching assistant give a summary of lecture at the board and then work out problems for the students. The teaching assistants go prepared with a small set of basic problems, simpler than those on homework, and illustrating points that have come up in lecture. However, the tutorials are also teaching assistant office hours, and students are encouraged to come with questions they may have. Any general guidance that the lecturer or the recitation instructors may have for the teaching assistants usually comes at our weekly staff meeting, held on Monday to set plans and directions for the week and beyond, but we typically leave the teaching assistants to come up with specific problems for their tutorials, perhaps in coordination with each other. The teaching assistants also take turns helping the lecturer generate the problem sets and solutions.

OCW: Please tell us about the role of the Common Room in the course. What was the impact of having a space where students could informally ask questions and work alongside each other on course assignments?

George Verghese: I think the evening Common Room is one of the best elements of the class, for those students—around a third to half of the class—who take advantage of it. I got the idea for it many years ago when visiting another university campus center after dinner, and found clusters of students sitting at desks and working collaboratively on homework and projects, though with no instructors in sight. For 6.011 Signals, Systems and Inference, we reserve a classroom for the three or four evenings that precede the day homework is due, and guarantee that at least one of the staff will be present there for 1.5-2 hours; usually we have the lecturer or a recitation instructor, as well as a teaching assistant.

“Our staff invariably finds the Common Room to be the most rewarding of the various settings in which they interact with students.”—GEORGE VERGHESE

We find students working individually as well as collaboratively, and periodically interacting with the staff, either at the board or at their desk—very immersed and engaged in the homework problems, and in sorting out ideas and misconceptions related to these. The staff will typically respond to student questions with other (well chosen!) questions or hints that guide them along, rather than with answers—and that makes for a very fruitful dynamic. We have never found the Common Room misused as a place to come and get fellow students to feed one solutions to the homework. I would absolutely recommend this to other faculty, if they have the staff resources and time. Our staff invariably finds the Common Room to be the most rewarding of the various settings in which they interact with students, and it is where they get to know their students best.


You can read more of Professor Verghese’s thoughts about teaching 6.011 on the Instructor Insights page of this course.

Keep learning! The following courses and Instructor Insights may be of interest to you:

More OCW Courses Offered by Professors Verghese and Oppenheim

Artist's depiction of the Cassini spacecraft, with Saturn in the foreground and a dark blue, starry background.Introduction to EECS II: Digital Communication Systems

An introduction to several fundamental ideas in electrical engineering and computer science, using digital communication systems as the vehicle. The three parts of the course—bits, signals, and packets—cover three corresponding layers of abstraction that form the basis of communication systems like the Internet.

The course teaches ideas that are useful in other parts of EECS: abstraction, probabilistic analysis, superposition, time and frequency-domain representations, system design principles and trade-offs, and centralized and distributed algorithms. The course emphasizes connections between theoretical concepts and practice using programming tasks and some experiments with real-world communication channels.

An audio compact disc.Discrete-Time Signal Processing

This class addresses the representation, analysis, and design of discrete time signals and systems. The major concepts covered include: Discrete-time processing of continuous-time signals; decimation, interpolation, and sampling rate conversion; flowgraph structures for DT systems; time-and frequency-domain design techniques for recursive (IIR) and non-recursive (FIR) filters; linear prediction; discrete Fourier transform, FFT algorithm; short-time Fourier analysis and filter banks; multirate techniques; Hilbert transforms; Cepstral analysis and various applications.

More about Communication, Control, and Signal Processing

6-003f11-th.jpgSignals and Systems

This course, a prerequisite for course 6.011, covers the fundamentals of signal and system analysis, focusing on representations of discrete-time and continuous-time signals (singularity functions, complex exponentials and geometrics, Fourier representations, Laplace and Z transforms, sampling) and representations of linear, time-invariant systems (difference and differential equations, block diagrams, system functions, poles and zeros, convolution, impulse and step responses, frequency responses). Applications are drawn broadly from engineering and physics, including feedback and control, communications, and signal processing.

2-14s14-thAnalysis and Design of Feedback Control Systems

This course develops the fundamentals of feedback control using linear transfer function system models. Topics covered include analysis in time and frequency domains; design in the s-plane (root locus) and in the frequency domain (loop shaping); describing functions for stability of certain non-linear systems; extension to state variable systems and multivariable control with observers; discrete and digital hybrid systems and use of z-plane design. Students will complete an extended design case study.

2-161f08-thSignal Processing: Continuous and Discrete

This course provides a solid theoretical foundation for the analysis and processing of experimental data, and real-time experimental control methods. Topics covered include spectral analysis, filter design, system identification, and simulation in continuous and discrete-time domains. The emphasis is on practical problems with laboratory exercises.

More on Assessment and Grading

6-01scs11-thIntroduction to Electrical Engineering and Computer Science I

Professor Dennis Freeman reflects on the advantages and limitations of using oral exams to assess student learning, and considers how online exams might help instructors offer scalable assessments that are personalized and productive.

6-034f10-thArtificial Intelligence

Teaching assistants Jessica Noss and Dylan Holmes describe the unusual grading system for this course, in which the final exam is optional, with each of its sections serving as a make-up exam for one of the course’s regular quizzes.

18-821s13-thProject Laboratory in Mathematics

Project Laboratory in Mathematics is designed to give students a sense of what it’s like to do mathematical research. In the Grading section of this course, Professor Haynes Miller and Susan Ruff describe their approach to grading and their experiences in developing (and revising!) grading rubrics.

Find insights like these on many other teaching approaches at our Educator Portal.