This weekend, Janurary 23rd, was the PiWars Conference 2021 - a robotics conference aimed at aspiring roboticists of all ages and levels of skill that runs alongside their yearly PiWars robotics challenge.
I was lucky enough to get the opportunity to present a 20 minute pre-recorded talk on some of the challenges I'd faced in the development of a visual SLAM system for the robot we had built for the 2020 competition.
It was the first time that I'd really had a crack at putting together a long-form video on any technical subject, as I'm rather used to either performing quick, snappy, relatively well-rehersed talks or more ad-libbed long-form in-person presentations on specific subject matters. Despite this, I had a fantastic time putting together the video and quite enjoyed the luxuries of being able to assemble and use detailed diagrams - the time required to do so, however, was not quite as appealing....
Well, it's certainly been a while since my last blog post, eh? But I have not sat idly by watching the sands of time pass me by! Quite the opposite! The last two years have been busy -
I've finished my bachelor's degree in Artificial Intelligence and Computer Science at the University of Sheffield, and for some mad reason I've decided to stick around, choosing to
start a PhD candidacy there, where I'm studying the information content retrievable by insect eyes....
Well, it's been a little longer than the "few weeks" I said that I'd take to test my little lucid dream inducer. But at last I am here to divulge the results!
Sadly, as you may have guessed already at this point, my hypothesis did not stand. Try as I might, not once did the device give me any better luck in lucid dream induction. However, I did learn
a few useful lessons in design as a result, as well as gain some insight into the way we react to external stimuli while asleep.
The Lessons
The first thing I learned - and very early on - is that it is very easy to be uncomfortable while trying to get to sleep. I wore the device without any problems for two weeks before attempting
to sleep with it on, during this time I barely noticed it; it fit extremely well under both jeans and thinner trousers and nobody else seemed to notice it was there either (something that I was
extremely relieved to discover - as I think it would have been fairly easy to confuse it with a geotag and assume that I had an azbo)
from the leg strap....
Following my last performance in the oxford heats of FameLab - a science communication competition - I returned this Wednesday, March 8th, to the regional final at the
Bullingdon, Oxford.
This time around I was attempting to explain, in my given 3 minute block, how artificial neural networks work. The whole event was being streamed, and you can see my segment above. The entire show
can be found here. It's worth the watch for a few hours of good, informative viewing.
Sadly (though not entirely without a small sigh of relief) this is where I get off the FameLab train, as I did not get through to the next round. Congratulations to Anna and Tomasz for their
winning performances!
This is certainly not the end for me talking about science though, as the Science Oxford and FameLab teams did such a fantastic job at introducing me to the world of science communication,
and I feel that it is definitely something I'd like to do more of going forward, so watch this space!
For my first project post of 2017, I think I'll about something a little off-the-wall.
So, without further ado, it's time for me to talk about something I've found fascinating since 2012 when I started dabbling in psychology,
hypnosis and mentalism: Lucid dreaming. A lucid dream is like a normal dream, but with the key difference
being that the dreamer is aware that they are dreaming. Once the dreamer has realised this fact, it is said that they can then exact a level of control
over the dream, which - according to some people - has some fantastic applications in getting creative inspiration to problems, helping
people get over nightmares, and encouraging you to actually sleep regular hours....
Last Thursday, February 9th, at the recommendation of my family, I rather impulsively entered FameLab UK - a
competition with the goal of finding "scientists and engineers with a flair for communicating with public audiences".
It was a genuinely fantastic night, and I really enjoyed meeting all of the people involved and performing, it was very invigorating! It's definitely the sort of thing I'll be trying to do
more of. Not least because I actually managed to make it through to the next round, so in a month I'll be doing it all over again talking about something else! I just hope I can think of something
to talk about!
Last month, I took a trip up to my educational mainstay - Sheffield University - in order to take part in the 3rd game jam run by the
Computer Science Society, a 30 hour games programming jam dubbed "ShefJam". Being the first year I'd actually
competed in the jam without having to worry about helping organise it, I was totally hyped.
This time around I had decided to put together a 3D game and focus on the graphics, something that I figured would be a lot more relaxing
than desperately scrambling to write a game from the base up again.
After faffing around for a few hours at the start talking to friends I hadn't seen in a while and failing to come up with any decent ideas,
I eventually wound up in a team with three friends of mine; Rebecca, Chris and Fred - though everyone just calls him "Guru". ...
On the back of the last post, I'm going with some slightly lighter reading. A lot of people have been asking about the background on this website,
so this post will go into detail about how it came to be and some interesting things learned.
Initially I had wanted to keep the background of my site as clean as possible, with a simple repeating background image. So I headed over to the infinitely
useful website subtlePatterns.com and started idly browsing through different patterns until I eventually came
across a very nice pattern called footer lodyas that kind of reminded me of the robotic-like stuff
from that movie 'The Signal'....
About three months ago I held a small talk on the benifits of using genetic algorithms to help debug and test code, as well as showing what I think is a fantastic
example of how critical data density can be to a system. All of this stemmed from a fairly simple task: Writing a program to play a game of Solitaire.
At the request of people who were not able to attend, this post will go through most of what was covered at the talk.
The task was set by our lecturer as an exercise in functional programming, so all the code referenced below will be in Haskell and the actual functionality of the system
will be kept intentionally vague.
The task was split out into two sections. First, we were instructed to implement the mechanics of a game of Eight-Off Solitaire
(the Wikipedia article on the game explains it's mechanics quite well), we were then asked to implement a basic AI
to play the game through to completion. The most logical way to approach this task - and the way many people did - was to work on those two aspects one after the other;
once the game was made, an AI could be made that could play the game well....
So, I've finally moved into the era of blogging! For those of you who know me in real life, then soon you'll find posts about everything you would talk to me about in real life right here in an easy-to-read web format! For anyone who happens to have wandered in from the internet, then you should hope to see posts covering a range of tech-related topics, from AIs to microcontrollers.
The blog will be updated as and when I come across something I think readers would be interested in, so feel free to add it to your favourites and check back every so often.
