Research Statistics: Fall 2020


Well. What a semester. It’s hard to adequately express how wild it has been to be a PhD student in the middle of the various social crises that have unfolded over the past nine months. A significant portion of the energy that I normally use for research has been diverted to some fundamental questions. Who does my research serve? What can I do to ensure that the things that I’m working on don’t simply exacerbate economic and racial divides? How can I justify spending this time and energy on something so distanced from the problems directly in front of me?

Robotics has not, historically speaking, been a field that addresses social need. The primary “consumers” of robotics research are large manufacturing corporations and irresponsible self-driving car companies. The only significant consumer-grade robot is the Roomba, a fantastically simple device that fails to do its job much of the time1. My hope that my research might buck this trend is a pale flame in the face of the headwind of the field’s history, and one that requires regular maintanence. It’s an abstract hope, too: If this goes right, and this other thing works, and I can manage to build this final thing, then maybe I’m doing work that matters. Day-to-day, my job is to read papers, do math, and code. Who can focus on these things when the world screams at you to justify yourself?

I’m saying all of this to try to get across the headspace that I’ve been in. I don’t mean to suggest that I am experiencing anything close to the full brunt of the pandemic; I have friends who have lost their jobs and others who have lost loved ones. Inequity is the story of the pandemic, and the larger story of contemporary America. What I’m trying to say here is that these are the issues that I’ve been grappling with. That mental and emotional work isn’t reflected in the statistics that I’m about to present, but it’s the bulk of the actual lived experience that these statistics represent. When I tell you something like “I read significantly fewer papers this past semester,” you need to understand that that energy didn’t disappear. It went into doing the work of understanding aspects of our society that I hadn’t had to look at face-on before. I occupy a privileged position as a PhD student in an engineering field, and that implies a debt to society that I intend to fulfill.

Fall 2020

In case you haven’t read the first blog that I did about my research statistics, you’ll probably want to read it here first. I’m going to elide discussion of my precise log setup here, so check out that post if you’re curious.

I’m not presenting research statistics for the summer because it was incredibly unproductive, and because I spent most of it revising a paper for resubmission and co-writing a grant application. Those processes were useful because they allowed me to revisit my research from a new perspective and plan some next steps, but I am also somewhat tired of writing papers that get repeatedly rejected. I’m sure that every academic has had solid research rejected, but it’s quite demoralizing now that I’m in the third year of my PhD. In any case, the fall semester was a sort of protected cove from those worries. I spent most of it trying to figure out how to be productive working from home, and succeeded only partially. More time than I would have liked was taken up by classwork this semester, but with the help of my friend Isaac2 I managed to figure out a new strategy for managing my research code which revitalized my work. We’ll write it up one of these days, so stay tuned.

The aggregate statistics for this semester are, incredibly, consistent with previous statistics: 27 days off and an average of 1.98 work items per day. The number of meetings that I had, however, almost doubled from Fall 2019, which is amazing since I didn’t even TA a class this semester. I can only assume that this is due to the low activation energy of scheduling a Zoom meeting. As I hinted earlier, I also read significantly fewer papers this year—exactly half as many.

Fall 2020 Plots

Fall 2020 trends plot Fall 2020 stacked dot plot Fall 2020 github contribution plot

The graphs for Fall 2020 reveal one interesting trend: smoothed total units of work peaked in October, which is about the time that I took on an additional writing assignment and I had a significant amount of classwork assigned. The fall-off in productivity after October could be attributed to burn out, but I also had to terminate a relationship during this time due to disagreements around CoVID. More than anything else, I have learned how intimately tied productivity is to my emotional state over the past few months.

I’ve included my Github contribution plot for this year as well in order to drive this point home. While I tend to record research coding as a single entry in my log, in reality from October to early December I coded like a madman. This is great for my research, as it means that I’m finally back to pushing it forward, but it represents a sort of manic turn away from negative emotions that I was not dealing with at the time. Those bright green areas represent pointless overwork, and often sleepless nights. This is the first time that I’ve had such hard evidence of my tendency to abuse work as a salve for my emotional wounds. Taking a few weeks off for the holidays has helped me right the ship, thankfully. I’m hopeful that going into 2021 I’ll be able to identify this trend early and keep myself from spiraling.

Fall 2019 Plots

Fall 2019 trends plot Fall 2019 stacked dot plot

I’ve also included the Fall 2019 plots here for comparison, but really the context is so different in Fall 2020 that I’m not sure how fair a comparison this is. One thing that’s worth pointing out, however, is that my productivity rebounded after Thanksgiving break in 2019, but not in 2020. I’m absolutely positive that this is indicative of the next-level burn out that characterizes my experience of work in 2020. My resilience is at an all-time low, and I’m sure that I’m not alone in that.

Coda

Everything isn’t terrible. I have found alternate avenues to be creative and find joy in life, even when the things that have traditionally brought me joy are either impossible due to the pandemic or deprived of joy by burn out. To wit, I’ve been learning to draw over the past year. Please enjoy some of my favorite things that I’ve produced recently.

Art: Surreal tree Art: Still life Art: Water Art: Metal and transparent objects Art: Guitar Art: Monstera

Published by Cannon Lewis on January 04, 2021