Robots evolve more quickly and efficiently after a virtual mass-extinction modeled after real-life disasters such as the one that killed off the dinosaurs, according to computer science researchers at the University of Texas. Their research shows how simulations of mass extinctions promote novel features and abilities in surviving lineages.
Researchers Risto Miikkulainen and Joel Lehman created computer simulations by connecting neural networks to simulated robotic legs with the goal of evolving a robot that could walk smoothly and stably. They introduced random mutations so a wide range of features and abilities would result. After hundreds of generations in which a broad spectrum of robotic behaviors had evolved, the researchers mimicked a mass extinction by randomly killing off 90 percent of the development niches. They discovered the surviving lineages had the greatest potential to produce new behaviors, and better solutions to walking were evolved in simulations with mass extinctions.
The researchers think their work could be used to develop robots that are better able to overcome obstacles, as well as human-like game agents.
From University of Texas at Austin
View Full Article
Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA
No entries found