Artificially evolving social robots

Carl Zimmer has written a fascinating piece on a study that simulated the evolution of communication in artificially intelligent robots.

The robots were small and mobile, were controlled by artificial neural networks, and could send and receive signals via infrared.

A group of robots was put into an arena with a light-emitting ‘food source’ and a ‘poison source’. The robots could only tell the difference when they got close enough to see coloured paper that the ‘sources’ were resting on.

Robots gained a point when they found the ‘food’ and lost a point when they stumbled across the ‘poison’.

Crucially, the researchers could electronically ‘breed’ the robots to improve their neural networks, so they could compare how the offspring of different combinations of best and worst performing robots would behave.

Zimmer notes some of the interesting results:

Here, however, is where the experiment got really intriguing. Each robot wears a kind of belt that can glow, casting a blue light. The scientists now plugged the blue light into the robot circuitry. Its neural network could switch the light on and off, and it could detect blue light from other robots and change course accordingly. The scientists started the experiments all over again, with randomly wired robots that were either related or unrelated, and experienced selection as individuals or as colonies.

At first the robots just flashed their lights at random. But over time things changed. In the trials with relatives undergoing colony selection, twelve out of the twenty lines began to turn on the blue light when they reached the food. The light attracted the other robots, bringing them quickly to the food. The other eight lines evolved the opposite strategy. They turned blue when they hit the poison, and the other robots responded to the light by heading away.

Two separate communication systems had evolved, each benefiting the entire colony.

The researchers have made some Quicktime video available if you want to see this in action and the scientific paper is available as a pdf file.

The research was led by Prof Laurent Keller, a biologist who specialises in understanding the evolution of communication networks.

There’s much more about this fascinating experiment and the other surprising types of behaviour that emerged from it over in Zimmer’s article.

Link to article Carl Zimmer’s article ‘Evolving Robotspeak’.
pdf of scientific paper.

Post a Comment

Required fields are marked *


Get every new post delivered to your Inbox.

Join 2,584 other followers