discrete thoughts on control education
Controls education and research at UW has a nice history. Generally, you can find control-oriented graduate programs in various engineering departments, including aerospace, mechanical, electrical, and industrial/systems, and chemical engineering. However, control-system thinking is useful in any engineering setting. When I joined UW, this historic account was shared with me by the late Prof. Clark. It traces the history of how control research and education looked like in the 1960s; just a note that this is (way) before I was born. There is a huge gap between then and now of course, but some other efforts that I am aware of included the Robotics, Control, and Mechatronics group and seminar series (RCM) that was overseen mainly by faculty in AA, EE (subsequently ECE), ME, and CSE. The RCM seminar was a relatively large gathering- it was held every Friday 2:30-3:30 pm and most often had outside speakers. At some point, the group and their departments who were supporting the RCM group got dispersed; for example, CSE started its own robotics colloquium.
These days, control system “minded” faculty (whatever that means) often have a number of other interests; some of us work on optimization/ML/AI, some on aerospace autonomy/autonomous cars/robotics, some on power grids, some on synthetic biology. What brings this diverse group of researchers together is their interest in dynamic decision-making, system architecture, and performance and robustness of synthesized dynamic systems.
There are a lot of awesome resources online to learn about controls-related topics; some recommendations:
Brian Douglas’ youtube channel and mathworks videos
Steve Brunton’s youtube channel
Chris Lum’s youtube channel
I am often asked about systems/control courses to take at UW, particularly for new M.S. and Ph.D. students. One of the luxuries of being at UW is that you don’t have a shortage of good courses to take every quarter. If you want to do Ph.D. in control, the stronger your linear algebra, mathematical analysis, and coding would serve you well- why analysis? I think a course like undergraduate analysis (Rudin blue book) would make you become a more careful reader; even if you don’t want to do proofs, you have to read others theoretical results you read papers in control theory. It is good for you- of course, you can continue taking more math classes, but something at the level of Rudin blue book is a must for phd students in systems and control. For M.S. students, the sequence typically goes like: linear algebra (510), linear system theory (547), multivariable control (548), and estimation/filtering (549). Then you can of course take courses that you provided more depth in certain areas, such as flight control, spacecraft dynamics, machine learning/perception, etc. I think some of this sequence will be updated in the years to come- introducing data analysis, statistics, ML; may be part of the linear algebra curriculum and then updating linear systems with some data driven aspects.
The other important aspect of your graduate education, in my view, should revolve around writing, presenting, and collaborating with others. This is so important- I seriously think as engineers we all need to become better communicators. This is not trivial- a lot of us are bad communicators because of various reasons- not knowing what to say of course is one primary reason. But you can also be a bad communicator if you are disorganized, do not think about your audience, oversell, undersell, too much or too little ego, underappreciated insecurities, how you articulate a given idea … for me, these are all work in progress- the process never ends … the sooner you start the journey, the better!