At OffKai, Panel 1 Boids-based simulation of the Hamtaro Circle – bringing algorithms to the Wota Pit https://www.offkaiexpo.com/event/boids-based-simulation-of-the-hamtaro-circle-bringing-algorithms-to-the-wota-pit/

• Presented by
  • Lilybweebs (youtube.com/@lilybweebs)
  • RTX Chan (404cafe.club)
• Arrived 6m late, in time to see the audience running in a circle around the perimeter of the room chanting, mimicking the video on screen.
• Boids stands for “Bird Oids”
• Boids imitate flocking using a combination of only three behaviors: Seperation, alignment, and cohesion.
• A version of this presentation will be presented and recorded on Lilybweeb’s channel after the con.
• Finding neighbors
  • Reynolds: Find each boid within a specific distance and angle of sight.
  • Couzin: Each force has a varied radius, allowing for more varied behaviours.
• Separation
  • Each boid tries to avoid running into its neighbors, which can be done easily if you know where all the neighbors are.
  • The algorithm is more or less adding a force away from each the neighbors and summing those forces
• Alignment
  • Try to match velocity and alignment
• Cohesion
  • Try to go toward the center of the group
• Edge detection
  • Turn around when approaching the edges of the simulation
• It doesn’t look like a hamtaro circle, so the model needs to be expanded
• Wota boids need to congregate in the pit and form a circle.
• Add a crowd. Boids have a chance of joining the pit.
• Circling behaviour has been well studied. by Silverberg, Couzin, and Jung
• Used a “mediator” from Jung, by adding a boid with both a predation and attraction factor.
• Inner circle looked too perfect; it should be uneven and chaotic. this can be fixed by changing the predation and attraction factors for each boid
• For the crowd, decide which boids are wota boids or crowd boids. Crowd boids have everything disabled except separation forces.
• Crowd needs to avoid the center, so we need to enable the predation factor. the predation force can be slightly bigger for crowd boids to add a berth around wota boids.
• Crowd is too active, so we can adjust their separation factor down when they are a crowd boid.
• We can add crowd boids in the middle, to match how some people get stuck in the middle in the real hamtaro circle.
• We can make some boids become wota boids
• Dynamic circle, change the size of the circle based on how many people are in the circle.
• Godot is not built for simulations, but it can work for small simulations
• Unfortunately, you cannot really directly control how quickly the circle forms, which direction the circle flows, or how many end up in the center.
• Q&A
  • Would you consider running this in not Godot?
    • If people are actually interested in getting a version of this for 3d concerts, would make a C-based version of it. Would also add a way to change parameters at realtime
  • Would you open source any of this?
    • A C-version would be open source. The Godot version is a mess, so it would need some cleanup. It would also need a way to change parameters at runtime.
  • Why was there a focus on efficiency? All the videos are prerendered.
    • Found out that they were going to do a panel late, and Godot was just a easy way to record a video. Not bring able to run it in realtime was disappointing, because boid simulations should be able to.
  • (missed the question)
    • You could add an exhausting parameter to have boids leave and rejoin the center.
  • What made you think of wota boids at the aquarium?
    • Boids were originally made for birds, but now it is almost always used for fish. This kind of simulation is used a lot for fish.
  • Have you considered the physical excercize that this could be used for?
    • This should be mandatory in PE.
  • How much caffine did you have for this project?
    • This was all done in 4-5 hour sessions, which was surprisingly little.
  • Have you considered having smaller wota circles in bigger circles?
    • Adding more features could dramatically improve the simulation?
  • If not Godot, what would you recommend?
    • Just write it in C.
  • (missed the question)
  • Is there a hamtaro hackathon in our future? are you prepared for hamtaro circles become your entire personality?
    • it sounds really fun, but the song has been stuck in my head
  • Has any peer review been done on this research?
    • No, there was going to be a white paper but they didn’t get around to it.
  • When will be get a hamtaro circle simulation for vr chat?
    • I don’t want to do it but now that you mention it, I have to look into it.
  • Do you know why the predation and attraction factors converge into the boids circling?
    • Since the boids try to avoid the center, they get pushed away, but the attraction factor pulls them in. the alignment and cohesion then moves the circle around
  • Would you do this with physical robots?
    • No, I have worked in robotics and I would do all of this virtually.
  • Have you considered changing the scene to see which spaces are more or less conducive to hamtaro circles?
    • That would be really complicated, could be its own paper, and boids are probably too simple for that.
  • Why comic sans?
    • I really like it. It’s friendly and easy to read for people with dyslexia
  • Have you considered adding a leader boid?
    • Yes, I have.