My journey to presenting research in a lecture theatre in Copenhagen began a year and a half ago around February 2023. I’d made it through the initial literature review of my undergraduate masters (MEng) technical project and was beginning to form the outline of my final dissertation. My supervisor (Chanelle Lee) told me I was exploring some genuinely novel aspects, and I might be able to get a conference paper out of it if I wanted to try at a later date. After graduation I asked her whether the offer would still be there to help me turn my technical project into a conference paper and she was more than happy to help.
I think there were a few things that interested me about giving it a go. The main thing is that I was interested in doing a PhD, but I wasn’t sure whether not it was for me and saw this as a perfect way to test it out. I also thought it would be a great development opportunity to bridge the gap from undergraduate to published researcher.
Two weeks later I was back in her office speaking about my project. We set out the structure of the paper, making sure it was clear what my research questions were, and how I was going to answer them. We also picked the conference to aim for: A[rtificail]LIFE.
The paper was in the field of swarm robotics, exploring a strategy to mitigate faulty agents in the collective perception problem. Basically, you’ve got a bunch of robots trying to figure something out, but each of them has only a limited view of their environment. How do you get them to effectively reach a consensus on some feature in the environment?
Each of them can make their own mind up based on what they perceive, and hope that the patch they’re exploring is representative of the whole, but more effective (usually) is for the robots to communicate with their neighbours. That way evidence can be passed across the robot swarm, across the environment, and hopefully this leads to reach a consensus more quickly, and more accurately. Essentially, robots alone: weak. Robots together: strong!
Robots working together can be less effective if some of the robots are faulty or malicious and communicate unhelpful information. My paper was presenting a strategy to mitigate against this unhelpful communication, while keeping the benefits of social communication. I tested the strategy using a simulation environment of robots that I wrote in the coding language Python.
After tweaking some code that I’d written during my technical project and outlining the paper I began to run simulations and write. One thing writing helped massively with was improving my ability to construct an argument. I met with Chanelle every two weeks or so via Zoom. I’d take her through what I was currently working on, and she’d ask questions and probe to make sure I was still on track and had considered multiple possibilities.
I ended up doing most of the work in bursts, between a couple of part-time jobs, and applying for a full-time job that felt a bit more permanent. I found it was easier this way as I could gain momentum, work through any issues that arose more quickly, and just generally keep my focus. Not being a current student meant I was a little out of practice and it was a lot easier to get distracted, and the first half hour or so of any work session would just be reminding myself where I left off. I eventually managed to get everything done with the help of Chanelle, and also Jonathan Lawry, another lecturer from Eng Maths who took a look over the paper before submitting. There was no guarantee the paper would be accepted, but a few weeks later I got an email through with the good news, along with some comments on things to fix before publishing.
It was an amazing feeling sending off the final version of the paper. All that was left was booking flights and accommodation, and a couple of months later I was in Copenhagen for the week with Chanelle, Jonathan, part of a surprisingly large Bristol contingency.
My presentation was on the Wednesday. With a full time job and the pull of a sunny weekend in Copenhagen beforehand – if I’m being completely honest – I wasn’t fully prepared, so I was quite nervous when I began. There was one man asleep in the front row, but I’m pretty sure he dozed off before I started, so I didn’t take it too personally. I made it through and had a couple of good questions. All in all a success!
ALIFE is a fun conference. There was an art exhibition, talks on subjects ranging from biophysics to computational sociology, and DJ performances from Lil Data and DJ Algorhythm. The food was good and everyone I met was friendly. There was a generally positive spirit and feeling of good will throughout the entirety of the conference.
I still haven’t decided if or when I want to do a PhD, but this experience gave me a lot more information, and what felt like a real taste of academic research.
Access the paper:
Patrick Shortall, Chanelle Lee, Jonathan Lawry; July 22–26, 2024. “A Macrophage-Inspired Approach To Fault Reaction In The Collective Perception Problem.” Proceedings of the ALIFE 2024: Proceedings of the 2024 Artificial Life Conference. ALIFE 2024: Proceedings of the 2024 Artificial Life Conference. Online. (pp. 3). ASME. https://doi.org/10.1162/isal_a_00711
Leave a Reply