Join the lab

From RobotLab

This is an open letter describing some of my expectations for undergraduate and graduate students working on a research project in the lab.

Expectations for undergraduate students

Although I do not typically "hire" undergraduates for work or internships, I would love to get you involved. There are many undergraduate research programs sponsored by Dartmouth, including the Presidential scholars programs (for sophomores), E.E. Just fellowships (for minorities), WISP internships (for women), as well as senior thesis opportunities or REU grants. You can work on existing projects, or create your own.

The typical undergraduate in our lab has skills in mathematics, computer science, or engineering, but we also have projects where artistic talents (particularly in computer modeling) would be particularly welcome.

You will need to make a time commitment of at least 10 hours per week. Purchase a research notebook (a standard composition book will be fine), and record your hours in that book each week. Also make sure that you write 1-2 pages in the notebook per week about ideas you've had, approaches you've tried, related work you've found, and sources where you've ordered parts or equipment. Expect to meet with me at least once per week for 20-30 minutes to discuss your progress. I also expect undergraduate students to enroll in the CS 54 course when it is offered.

Expectations for graduate students

If you are considering choosing me as your adviser for an M.S. or Ph.D. degree, talk to me as soon as possible so that we can figure out if our interests and expectations are a good fit.

• Courses. Robotics requires a broad background, not all of which is available within the C.S. department -- you should expect to take courses in mathematics (particularly real analysis and differential geometry), engineering (linear and non-linear control), and the introductory robotics class that I teach, C.S. 54.

• Time commitment. I expect the time commitment to be 40-50 hours of efficient work per week. The first term of graduate school can be difficult to adjust to, so the expectation is that about 10-15 hours per week will be spent on research work during this time. For the remainder of the first year, about half of the available time should be devoted to research, with the remaining half devoted to classwork. From the second year, students are expected to be doing full-time research. I expect a Ph.D to take slightly less than six years, and a M.S. to take slightly less than two years, although individual progress may be slower or faster.

• Measuring progress. Research progress can be hard to measure, at least locally. A useful metric is how many pages you write each week in your notebook. For a grad student doing full-time research, a good target is 4 pages of figures, equations, data, and brainstormed ideas per day. That's 1000 pages each year, 1500 pages for a M.S. (half-time first year), or 5000 pages for a Ph.D. (half-time two years, total of six years). I'm very serious about this; most of my evaluation of how well you are doing will be based on this metric. Quality matters too, but I will assume you are doing your best!

• Adviser meetings. During your first year, we'll meet three hours per week, broken up into two meetings. You may feel that there's not enough time between meetings to get any work done. However, remember that these meetings should not be regarded as meetings with your supervisor, but with a collaborator. They are a great time to get some research done, and two people are often smarter than one! As you become more comfortable with the research process, we'll meet less and less. From the second year onwards, we'll probably only meet one hour per week, unless we are working particularly closely on a problem.