Instructor

Shashikanth Suryanarayanan, Abhishek Gupta

Semester

Autumn ‘20

Course Difficulty

Not very difficult for someone with a reasonable appetite for math.

Time Commitment Required

Reasonable efforts were required to be put in outside class hours.

Grading Policy and Statistics

Quite unremarkable.
14 AAs, 21 ABs, 38 BBs out of 168 students. On the positive side, only 2 FR grades were awarded.

Attendance Policy

Not compulsory

Pre-requisites

Considerable mastery over the content of MA108 (ODE theory, Laplace Transforms) is almost essential to ace the course, although the professors may cover things at a slow enough pace that it is easy to understand from scratch.

Evaluation Scheme

20% Weekly Tutorials (8 total)
15% Best 2 out of 3 Quizzes
25% Midsem
40% Endsem

Topics Covered in the Course

Block Diagrams, Feedback and feed-forward systems, Adaptive cruise control (ACC), Fourier and Laplace Transforms, LTI, BIBO Stability, LCC ODEs, Linear control design, Gain and Phase Margin

Teaching Style

The entire course was taken by Prof. Shashikanth. Prof. Abhishek would be around in the lectures and would occasionally conduct doubt sessions, but did not teach anything as such.
Teaching was reasonably slow and covered real-life applications of the topics. The class interaction was quite healthy. While slides were provided, they were typically taciturn and one would have to also study from class notes in order to do well in the exams.

Tutorials/Assignments/Projects

Weekly Tutorials - Probably the most significant component of the course evaluation. A tutorial session was held every other week, but from my experience, it was quite futile. Questions were incredibly lengthy and often very tedious. The professors make all of the questions on their own and it isn’t even remotely close to anything you’ll ever find online. Most of the problems are design or simulation based and heavily involves MATLAB. For someone with no functional knowledge of MATLAB and a solid understanding of the content covered in class, the tutorials would be a really arduous task.

Feedback on Exams

Quizzes were typically easy and short and a mix of objective and subjective type questions. Easily the most scoring component of the entire course.
The midsem and endsem were quite lengthy and resembled the tutorials to a great extent. There were several design-based problems, but in the absence of MATLAB, one would often have to indulge in dreadfully long calculations to obtain an estimate of the answer. Several questions in the paper carried a very high weightage (17-20 in an 80 mark paper) and consisted of tens of sub-questions one following another. So, if someone made a mistake in one of the earlier sub-parts, they would end up losing a large fraction of their grade.
The exams and tutorials weren’t evaluated very leniently and class averages were not very high either.

Course Highlights

The last part of the course on Linear Controller Design is extremely interesting, irrespective of whether you really like the overall theme. As Prof. Shashikanth said (who happens to be one of MIT’s top global under 35 innovators, read up on SEDEMAC, he’s the founder), most of those problems come straight from the automotive industry. Even the initial tutorial problems, daunting as they seem, are very realistic and you will eventually appreciate the effort put into designing them.

Course Importance

Provides a glimpse of the Controls field, in a way what SysCon does in a nutshell. If you liked the field, you can also consider picking up some Controls electives from SysCon.

References Used

Surprisingly, very little of the material is even remotely traceable online.
For the parts on ODE theory and LTI systems, “Signals and Systems” by Oppenheim is recommended, but I didn’t find it overlapping much with the course content.

Other Remarks

Attend classes, or at least try to regularly catch up on the recordings! This is one of the more demanding courses.

ME 311 Review By: Aditya Iyengar