GRIP, Internships Abroad Research at NUS

GRIP: Engineering Research in Singapore

The project that I worked on over my internship tackled the task of Multi-Agent Pathfinding. For instance, imagine that many robots are working in a large warehouse, where each robot needs to reach their individual destination to say, deliver a package. How do we map and calculate the best paths of each robot such that the complexity of the algorithm is relatively low, but also creates a solution where each robot gets to their destination relatively fast? In MAPF in MARMoT lab, we used Reinforcement Learning as a basis to our solution.

The lab was structured in a very collaborative way. Every week, the professor would organize lab-wide meetings and lunches. During the meetings, each member shared progress on their research in a presentation format, and the professor and other peers would give probing questions asking about the results and methods. In turn, the presenter would have to defend their ideas and process. At the end of the meeting, each researcher was given a new and stronger direction to continue doing research in for the next week. Within the lab, there were many different departments, and students from different departments normally do not interact with one another, but these lab-wide meetings allowed researchers to hear about new ideas from various other departments. As a result, some of these meetings led to intensive discussion and were a lot of fun to attend. After these meetings, we would have lunch together at the dining hall at NUS, and there we were able to have more informal conversations and bond between the lab members.

This internship experience has changed the way that I view graduate student life. Previously, I assumed that research was similar to the the undergraduate environment of taking classes and to excel at your studies indicated a propensity to enter academia. However, I found this to not be true. Research at the PhD level requires a lot of self-regulation; projects are not structured in the way that assignments are. Advisors are generally much more hands off, and the work that you do is more "opinionated" and the directions you take are dictated by you. You create your own timelines, and advisors are there to give recommendations but not hard deadlines. I think that there is no substitute for knowing about whether or not research is the correct path for me besides attempting it in the first place. This summer was a pretty nice way to do that.

Finally, it was lovely to get a closer bond with many of the PhD students working in the lab. At the end of the program, our professor threw a small pizza party gathering, and I was able to spend a lot of time meeting some of the researchers I hadn't yet talked to, as well as make more personal conversation with my mentor in the lab. Overall, the summer allowed me to learn a lot about research as well as create many lasting friendships, and felt very fulfilling.