 |
Author |
Topic: Modelling of Artificial Life --- Help Needed!!! (Read 3231 times) |
|
mikky_m
|
Modelling of Artificial Life --- Help Needed!!!
« on: Sep 30th, 2004, 3:50pm » |
|
Hi I am about to start writing a paper on the topic of modelling Artificial Life, and I hope to use Processing as a modelling environment. So I would appreciate any ideas, questions, statemants, theories, etc on the topic. It would be very much appreciated. Thanks in advance..
Mike
|
|
|
|
|
TomC
|
Re: Modelling of Artificial Life --- Help Needed!!
« Reply #1 on: Sep 30th, 2004, 4:10pm » |
|
It's a big area - behavioural robotics, cellular automata, l-systems, neural networks, genetic algorithms, swarms, flocking, and more all come under some definition or other of artificial life.
What kind of systems are you interested in?
Are you looking at learning about real organisms from studying artificial ones, or building a realistic virtual creature, or solving an engineering problem with ideas from nature? Or something else?
|
|
|
|
|
mikky_m
|
Re: Modelling of Artificial Life --- Help Needed!!
« Reply #2 on: Oct 1st, 2004, 12:01am » |
|
Hi Tom first of all thanks a lot for replying. Cellular automata is something that I am particularly interested in. And I would say as you mentioned it solving engineering problems using ideas from nature is of a great relavance. Engineering obviously meaning not bridges and etc, well I am sure you know what I mean. So basically I am looking for insperational ideas and just useful thoughts which I can further develope in my research. Plus I wanted to use Processing as to demonstrate possible algorithms, solutions and problems. So any thoughts on that would be most exciting. Just touch on points, as I will research into it in detail at the next stage of my to be work. Thanks a lot again. Mike
|
|
|
|
|
mikky_m
|
Re: Modelling of Artificial Life --- Help Needed!!
« Reply #3 on: Oct 1st, 2004, 12:08am » |
|
The clever word I was looking for would be Biomimetics. 
|
|
|
|
|
TomC
|
Re: Modelling of Artificial Life --- Help Needed!!
« Reply #4 on: Oct 1st, 2004, 12:42am » |
|
I've worked on a couple of bio-inspired systems before, so I'm keen to talk about it, but not this openly. At least narrow it down a bit!
Here are some things I'm generally inspired by:
Craig Reynolds' flocking and steering behaviours (have a look at mflux's Processing implementation too)
http://www.red3d.com/cwr/boids
Karl Sims' work on evolved organisms:
http://www.genarts.com/karl/evolved-virtual-creatures.html (what's he doing now?)
From the DEMO lab at Brandeis:
The Golem project:
http://demo.cs.brandeis.edu/golem/ (evil creepy creatures evolved and then rapid-prototyped)
Evolved lego:
http://helen.cs-i.brandeis.edu/pr/buildable/crane/ (crane and emergent cantilever bridges)
Evolved 3D L-system:
http://www.demo.cs.brandeis.edu/pr/evo_design/evo_design.html (coffee tables)
Also, you might like to take a look at John Frazer's An Evolutionary Architecture and The Algorithmic Beauty of Plants which are both now free to download (both pointed out here recently by amoeba, I think).
http://www.aaschool.ac.uk/publications/ea/intro.html
http://algorithmicbotany.org/documents/
A couple more inspirational books are Turtles Termites and Traffic Jams by Mitchel Resnick and Vehicles: Experiments in Synthetic Psychology by Valentino Braitenburg.
If you're interested in a review of academic research, you could do a lot worse than checking out "Biologically-Inspired Computing Approaches to Cognitive Systems : a partial tour of the literature" by Dave Cliff of HP Labs: http://www.hpl.hp.com/techreports/2003/HPL-2003-11.html
If you want more, you're going to have to narrow it down
|
| « Last Edit: Oct 1st, 2004, 12:48am by TomC » |
|
|
|
|
Fish-face
   
|
Re: Modelling of Artificial Life --- Help Needed!!
« Reply #5 on: Oct 1st, 2004, 8:23am » |
|
Artificial Life is a wonderfully broad topic. Emergent behaviour is one area that fascinates me - that a small system with simple rules can produce fantastically complex and beautiful results. This is demonstrated in the particle systems that flew around here a while ago.
I also created a flocking and shoaling system based on this - another form of emergent behaviour observed frequently in nature.
I'm currently working on a neural net in processing, and it's definitely fascinating. When it's nearer completion I'll post a more detailed review of it, but it's having trouble learning at the moment.
If you can get something to evolve an actual moving creature, we'd all be interested - there was a lot of talk about that a while ago, with the possibility of using a sodaplay style setup, but we had trouble creating a genetic code.
One of the most important things in evolutionary systems is the evaluation of the mutations - you need to fine tune the mutation and reward feedback quite carefully to end up with a stable, good result. If you can demonstrate an algorithm for evaluating, say, the success of an evolved structure then that would be very cool.
Cellular automata is amazing - because again it is so simple, resulting in such diverse results. It's almost trivial to create a set of rules for a good cellular automata, but who knows what the output will be. A good idea might be to use cellular automata to model something seen in real life - people here have constructed simulations of sand, imiscible liquids, snow and all sorts. If I ever get round to it, I'd like to do a Reaction-Diffusion system based on CA, but I'm not too strong on the maths side of it.
Hope I've given you some ideas!
|
--
The C@ S@ on the M@
=Fish-Face=
~#Chris#~
|
|
|
|
