My name is Adam Wermus and I am, at the time of writing this, 26 years old living in Toronto for a co-op position with SideFX. I went to James Madison University where I double majored in Physics and Math, and minored in Music, Astronomy, Business, and Jazz. After graduation, I played trumpet at Busch Gardens for the rest of 2012 and then for Royal Caribbean in 2013. I then went to the University of Bradford in England to get a masters in Computer Animation where I also joined a gymnastics team and found a British Style Brass Bands to play cornet with. In 2014, I had the opportunity to go to Brigham Young University to do research(more on that in a moment) with Professor Seth Holladay. I completed my master’s thesis work there graduating from the University of Bradford in July of 2015 but staying at BYU that whole time for research and programming classes. At the end of the summer attended Siggraph. It was there that I got in contact with people from SideFX that led to this 3 month co-op opportunity. It goes from September-December.
At James Madison University, two of my math professors developed-on accident- a new method to solve Ordinary Differential Equations. This is a powerful tool that can convert any system of ODEs into a Maclaurin series solution. This means in terms of accuracy it can beat the Euler and Runge-Kutta methods making it worthwhile to test on simulations in computer graphics. Plus I feel like a kid in candy land taking what I learned from Physics and Math and putting it into animation.
One night when I was still in England(early 2014) I was dreaming/hoping/praying for an idea that could be my big contribution to the animation industry-that idea that once I do enough to show its merit-would lead to an opportunity at a studio. One night I couldn’t sleep and then the Parker-Sochacki method came into my head and I couldn’t shake it(you know you’re a nerd when you can’t stop thinking about a math theory when you couldn’t sleep all night and it’s 6AM). I looked on Wikipedia and this quote stood out: “… making it convenient for high speed computations…” It glowed with me like I was onto something. But then I had doubts thinking-if this is a good idea then surely people in the computer graphics community would have picked it up by now. I then realized that math theories are not like Taylor Swift in that they don’t go viral.
I looked everywhere online and could not find a single research paper where this method was discussed in computer animation. I also e-mailed several experts and the consensus was that they haven’t heard of it BUT anything that can speed up simulations will make people happy.
My vision is to implement the Parker-Sochacki method to simulations in computer graphics. My idea is this: if the accuracy is higher than Runge-Kutta, can Parker-Sochacki take larger time steps-and because it is taking larger time steps-speed up simulation time? On the things I have tested it has been a yes. More on that soon.
Take Care! You are awesome!