Week 11: Simulations, Dynamics, and Artificial Life

Dr Charles Martin

Semester 2, 2020

positive affirmations

“I have started my major project.”

“I have committed my latest work to gitlab”

“I have been adding to my SoO as I go”

“I have looked at my MP at the test URL”

“When I need help, I ask on the forum”

MP: process & policies

the process is the same as the assignments

you must submit an interaction statement and an artist statement

no late submissions (so don’t leave it till the last minute!)

make sure you reference all code/assets/inspiration in your statement-of-originality.yml (and re-read the course academic integrity policy)

MP: getting help

you still have a chance to get help in your week 12 lab.

feel free to share your major project problems / solutions / ideas on the course forum. We love to see your questions and discussions.

there’s nothing wrong with showing off your MP if you’re comfortable.

major project tips

sketch!

it’s not too late to try lots of different angles on the theme

'user' testing

watch a friend use it (without telling them anything first)

what do they do? where do they get stuck?

what's the wow factor?

what is the thing which makes your major project stand out amongst all the others?

how will you make sure that your sketch presents itself as the best version of that “wow”?

write your SoO as you go

you don’t want to end up in an academic misconduct situation

push to GitLab all the time

rookie mistakes

not reading the spec

this is worth 50%—you should read (and re-read) the major project page

making a game

where’s the artwork in that? what are you trying to communicate?

would it make sense on the wall in a gallery?

not testing at the test url

see point #4 on the submission checklist

putting an example in the microwave

you might have gotten away with it for an assignment

not acceptable for your major project

conceiving of the theme too narrowly

think big, think broad, think outside the box

use your artist statement to tell me how you interpreted the theme

don’t make a powerpoint

code theory

inspiration from nature...

Movement, Friction, Force, Fields…

Growth, Reproduction, Decay…

Collaboration, Ecosystems…

These are all things that change. We want to simulate these systems, and explore dynamic and emergent behaviours.

Daniel Shiffman

The Nature of Code

The Nature of Code

Textbook for a second course in creative coding.

Available for free online.

And in “pay what you like” for a PDF.

p5.js “simulate” examples generally come from the book.

(N.B.: uses Processing not p5.js)

Vectors and Physics

Godfrey Kneller

Portrait of Isaac Newton (1642-1727)

1689

Simulating Motion

Simulating “realistic” motions by using simple rules of motion.

Easiest to do this in 2D!

We’re going to use p5.Vector, an object to wrap up an x,y value. (see ref)

var vec = createVector(10,25);
var x = vec.x;
var y = vec.y;

Ball Model Skeleton

var ball;

function setup() {
  createCanvas(400, 400);
  ball = {
    pos: createVector(78,231),
    vel: createVector(2,-10),
    
    drawBall: function() {
      fill(255,0,0);
      stroke(0);
      ellipse(this.pos.x, this.pos.y, 50, 50);
    },
    
    updateBallPos: function() {
    }
  }
}

Draw Loop

The draw loop is going to draw the ball. And then “update” the ball.

function draw() {
  background(220);
  ball.drawBall();
  ball.updateBallPos();
}

Moving Ball

Add the velocity to position at each frame.

This goes in updateBallPos():

this.pos.x += this.vel.x;
this.pos.y += this.vel.y;

I guess this means that the velocity is measured in “pixels per frame”.

Bouncing Ball

// check boundaries
if (this.pos.x > width || this.pos.x < 0) {
    this.vel.x *= -1;
}
if (this.pos.y > height || this.pos.y < 0) {
    this.vel.y *= -1;
}

Bouncing Ball with Gravity

Gravity is a constant acceleration towards the ground.

We can model gravity by constantly increasing the velocity in the y direction.

this.vel.y += 0.5;

Completed Bouncing Ball Object

(link)

  ball = {
    pos: createVector(78,231),
    vel: createVector(2,-10),
    drawBall: function() {
      fill(255,0,0);
      stroke(0);
      ellipse(this.pos.x,this.pos.y,50,50);
    },
    updateBallPos: function() {
      this.pos.x += this.vel.x;
      this.pos.y += this.vel.y;
      // check boundaries
      if (this.pos.x > width || this.pos.x < 0) {
        this.vel.x *= -1;
      }
      if (this.pos.y > height || this.pos.y < 0) {
        this.vel.y *= -1;
      }
      // apply gravity
      this.vel.y += 0.5;
    }
  }

Game of Life

We’ve simulated a ball.

Now how about simulating an ecosystem?

What’s the simplest way to do that?

Dith Pran/The New York Times

John Conway at Princeton University

1993

Conway’s Game of Life

A little game of life game

John Conway was a mathematician with many contributions including to “recreational mathematics”.

Game of Life is an example.

(Died of COVID-19, January 2020)

Cellular Automaton

Some Cells

  • Cells live on a grid

  • Each cell has a state (true or false)

  • Each cell has a neighbourhood (of adjacent cells)

  • Cells change state according to their neighbourhood

Rules of life

Some Cells

Each cell is updated in each frame.

If it’s alive and has more than three neighbours, it dies (overpopulation).

If it’s alive and it has less than two neighbours, it dies (loneliness).

If it’s dead but has exactly three neighbours, it comes to life! (birth).

In p5.js

  • We will need a 2D array of Boolean values.

  • need to draw the array on the screen at each frame

  • need to find out how many neighbours a cell has

  • need to update the array each frame according to the rules

Here’s an example.

Boids and Flocking

Let’s look at another classic artificial life concept..

Craig Reynolds

simulated boid flock avoiding obstacles

1986

Boids

a flock

A simulation of coordinated animal motions (e.g., birds, fish) by Craig Reynolds in 1986. (original paper)

Generic creatures are named boids, represented by little triangles.

Widely discussed as an example of artificial life that shows emergence of complex behaviour from simple rules.

Three simple steering behaviours: separation, alignment, and cohesion.

A Boid

a boid

Has a direction and a velocity.

Can steer left and right by angle.

Reacts to other boids in a small neighbourhood

Separation

Don’t crowd.

Adjust heading to avoid nearby boids.

Alignment

Move in the same direction

Calculate the average velocity vector of nearby boids and adjust our velocity to approach that.

Cohesion

Move to the center of the group

Calculate the average position of nearby boids and aim towards that.

Boids Demo

dan shiffman’s boids

charles’ boids

why simulate?

This is where coding multiplies your creativity.

Once you set up a system, you get the emergent behaviours for free.

Try doing that in Photoshop, Final Cut Pro, Premiere, or Ableton Live…

Creativity in Artificial Life

Creativity theorist Margaret Boden writes about three types of creativity.

  • novelty (newness)

  • utility (value)

  • surprise

Artificial life art is full of surprise. These systems shouldn’t be so compelling, but they just are.

Ref: Margaret A. Boden; Creativity and ALife. Artif Life 2015; 21 (3): 354–365. doi:10.1162/ARTL_a_00176

Artificial Life Art

These systems are interesting.

But are they creative?

What have other artists done with them?

Therese Schubert

Sound for Fungi. Homage to Indeterminacy (link)

2020

Therese Schubert

Always Dead and Alive (link)

2019

Juan Manuel Castro

Heliotropika (link)

2018

author: Gary Greenfield and Penousal Machado (various artists)

Ant- and Ant-Colony-Inspired ALife Visual Art. Artif Life 2015; 21 (3): 293–306. doi:10.1162/ARTL_a_00170

2015

further reading/watching

Creativity and ALife

Artificial Life: Art and Creativity (Journal Issue)

Art and Interaction Computing Bibliography

Questions