Senior Spring by Alex M. '21

Finally, finally, I’m in my last semester of undergrad. I need three more classes to graduate, exactly enough to be a full-time student, and I’m taking a fourth for fun and knowledge and hopefully preparing for grad school. All my classes are remote, and I don’t have campus access, so I can’t go to rowing practices. This means my UROP is my main organized extracurricular activity, and I’m hoping to spend a lot of time on it, especially since I’m taking a slightly lighter course load than usual, so I’m claiming this spring as a “research semester.” This is mostly an aspirational goal, because, you know, classes.

Without further ado, here are the details on my classes & UROP & life as a second-semester senior.

14.03: Intermediate Microeconomics

This is the last HASS⁠01 Humanities, Arts, and Social Sciences elective I need to graduate. It also is my only class this semester with problem sets. I like problem sets, so I’m fine with this, but I find it ironic that my technicals are all about writing papers and my HASS is just like “do these constrained optimization problems by Sunday.” 02 to be fair, we do also read economics papers + discuss them So far, I like the professor and the TA a lot, and I hope I keep liking them. 

I think econ is interesting, but I feel a little bad about counting 14.03 as one of my eight HASSes. I feel like I should have to do more reading and writing to receive HASS credit. Pre-COVID, I envisioned myself taking a bunch of three-hour seminar HASSes senior year, sitting around a table discussing big ideas and drinking coffee. This dream was dashed by my hatred of breakout rooms and my inability to focus on Zoom classes lasting longer than 50 minutes. So, 14.03 it is.

17.315: Health Policy

This is the last HASS I need for my public policy concentration, and it’s much HASS-ier than 14.03 (meaning it has readings, in-class discussions, and essays). I really like Prof. Campbell, who I also had for 17.30⁠03 Intro to Public Policy sophomore year, and I’m excited to have her as a professor again. I’m also looking forward to learning more about how and why the US healthcare system sucks, and discussing alternative healthcare models that would maybe suck less. 

In general, I really like taking public policy classes. I think it’s interesting and useful to learn about policies that have been proposed or tried before, and to learn about different mainstream schools of political thought. This seems like it should have made me more centrist,⁠04 at least based on the people I encounter in my policy classes (compared to my friends) lol but it most definitely has not. 05 although it has helped me have political conversations with centrists & libertarians & such I think it would be weird to use this platform to promote my political views,⁠06 i don't think they're too hard to guess though so I’ll just leave it at that.

16.343: Spacecraft and Aircraft Sensors and Instrumentation

This is a project-focused grad class on sensors: what are they, how do they work, and why might we use them? It’s taught by Prof. Kerri Cahoy, my UROP PI and an all-around great professor and great person, and by Dr. Becki Masterson, my ex-UROP PI and also a great professor and person. So, this class will probably be taught really well!⁠07 very exciting in the age of Zoom I also like that it’s project-focused, which fits nicely with my goal of having a research semester because class projects are basically research, right?

My main experience with sensors comes from my UROP. Last summer, I worked on an Extended Kalman Filter that turns measurements of Earth’s magnetic field, gyroscopic rate of a spacecraft, and the angle of the sun (using solar panels as coarse sun sensors) into knowledge of satellite orientation ⁠08 aka attitude relative to Earth. This was a cool introduction to sensor applications, but I mainly treated the actual sensors (a magnetometer, a gyroscope, and solar panels) as black boxes with bias and Gaussian noise. I don’t know how these sensors work internally, but I’d like to find out, which is one of the reasons I’m taking 16.343.

The main assignment in 16.343 is a very open-ended sensor project, culminating in a final paper and presentation. I’m currently brainstorming project ideas, which basically means I’ve been trying to think of new, cool sensor applications. I’m vaguely intrigued by doing a sensor design project instead, but I know so little about sensors that I’ll probably stick with applications for the project, and just try to pick up some sensor design concepts from the 16.343 lectures + readings instead.

16.83: Space Systems Engineering

16.83 is the last class I need to finish my major. It is also a CI-M⁠09 communication-intensive class in your major -- you need two to graduate, and they're usually lab or project classes , and half of the space-focused option for the course 16 undergrad capstone sequence. I took 16.831, the other half of the capstone sequence, seemingly ages ago, in spring 2020. 16.831 was chaotic ⁠10 the fact that I took it in spring 2020 was part (but not all) of the reason for the chaos but surprisingly fun, and it unironically changed my life — it got me interested in satellite research, and then I started UROPing on a satellite controls project as a continuation of my work in 16.831, and now I’m planning to go to grad school to study satellites and other space systems.

My hope is that 16.83 will be, like 16.831, surprisingly fun, if probably less life-changing. There are several promising factors at work: I like the teaching staff (which, happily, includes Prof. Cahoy), the class is pretty small but I still have a couple of friends in it, and most of the work will be on an open-ended space systems project. 11 as you can probably tell, I really love space systems Unlike 16.831, there isn’t a prearranged class project; together, the class will determine the space mission concept that we’ll work on. Regardless of how the brainstorming phase turns out, I expect that I’ll like the project, and I’ll take this as another win for my ~research semester~.

UROP

I’ve wanted to blog about my UROP for a while, but usually end up just UROPing instead as soon as I start trying to write about it. But anyway, when I took 16.831, the class project was a 3U ocean-observing CubeSat called BeaverCube. Due to COVID delays and also just normal delays, BeaverCube still hasn’t launched, and I’m still working on it. Specifically, I’m working on attitude ⁠12 aka orientation determination & controls, or basically figuring out the orientation of the satellite and then reorienting it to meet various mission goals, like pointing its cameras towards the ocean.

During the spring (in 16.831) and summer, I mostly focused on attitude determination, or determining BeaverCube’s orientation relative to the Earth. This is why I worked on the Extended Kalman Filter I wrote about earlier. Since the fall, I’ve been working more on controls, or pointing BeaverCube in the right direction. This has proven to be a much harder problem.

At a very high level, BeaverCube is difficult to control because its only actuators are magnetorquers, which torque the satellite perpendicular to the local magnetic field only, so torque is restricted to two dimensions. As BeaverCube orbits, Earth’s magnetic field changes, and so does the plane perpendicular to the magnetic field, which means the two dimensions of available torque change over time. So, theoretically, if you can predict the magnetic field over time, and plan a maneuver very carefully, you should be able to torque in all three dimensions, but it’s hard and computationally expensive to do this.

This is basically what I’m working on — optimizing control trajectories over time and then testing to see if they actually work. My grad student has written a bunch of code for this in MATLAB, and I spent most of the fall porting it to C++ and then optimizing it to run in reasonable time on a Raspberry Pi, because BeaverCube has a Raspberry Pi-based flight computer. Recently, I’ve spent a lot of time debugging my C++ code, both fixing differences between the C++ code and MATLAB code and troubleshooting more global problems with getting the code to work in realistic scenarios with noisy sensors and external torques.

This spring, I’ll be continuing my current debugging work and trying to get BeaverCube’s attitude determination & control software to a flight-ready state, as BeaverCube is now scheduled to launch this August. I’m excited to see things coming together, and I really hope I’ll be able to attend the launch this summer! I’m also working on a longer blog post about my UROP, so keep a lookout for more information about BeaverCube and how it convinced me to apply to grad school soon.

Life

Life in quarantine is objectively not very exciting, but time marches on anyway. Here are some small bits of joy from the last couple of weeks:

• My roommates have a dog, and she is perfect, and I pet her all the time. She is very soft. In fact, she is so soft that she sometimes gets compliments from strangers when they pet her. Here she is trying to pull an entire tree out of the ice at Fresh Pond:
  

  her name is sam. everyone say hi sam!

• My girlfriend and I made raspberry macarons for Valentine’s Day, and they were delicious:
  

  i wish i still had more

• We also made many other delicious things to eat, including but not limited to poke bowls:
  

  we used lox instead of actual sushi grade fish but it was still so good

• We also did a couple of snow hikes and many neighborhood walks. It’s been really nice to get outdoors and experience the real world, as opposed to Zoom world. It’s also been nice to start recognizing some landmarks around my apartment and adjust to actually living off-campus in a new area. ⁠13 not to dox myself but I live in {redacted neighborhood} in Somerville, which is not super close to MIT, and I haven't spent much time here before
• I did my taxes, which sounds boring, but it took under an hour, and I’m getting a big enough refund to almost cover one month of rent + a new dresser + a new desk chair. 14 I only paid $11 in federal income tax and got the rest back Hurray for my unexpectedly low 2020 earnings! 
• This is extremely embarrassing to admit, but I have been playing Clash of Clans a lot with my little brothers. 15 even more embarrassing: I started playing in like 2014 because my high school rowing team briefly had a clan, then essentially went on a several-year hiatus until last summer, so you could say I'm pretty serious Our clan is literally just the three of us, but we operate sixteen separate Clash of Clans accounts, and we totally kick ass. I’m pretty sure this is not allowed, so if anyone from Supercell is reading this, no you aren’t <3

All in all, things are going pretty well so far. It was weird to start the semester off-campus, but I actually really like writing my silly little MATLAB scripts in my cozy little apartment. But also, I hate Zoom and can’t wait to delete it forever. So it goes.

I hope that this semester, I actually do spend more time on research⁠16 aka my silly little MATLAB scripts, which are good and less time on classes.⁠17 aka Zoom, which is bad I also hope I get vaccinated by the summer, and I am ready and waiting on the standby list in case a freezer full of vaccine starts melting or MIT Medical accidentally receives 5,000 doses in their next shipment instead of 500 or something.

I’m cautiously optimistic about both of these goals, and about the fancy desk chair I’m probably going to buy with my tax refund, ⁠18 per Jeremy’s recommendation and about this semester in general. Here’s to hoping everything goes at least a little bit according to plan.