Home // COGNITIVE 2016, The Eighth International Conference on Advanced Cognitive Technologies and Applications // View article

Evolving a Facade-Servicing Quadrotor Ensamble

Authors:
Sebastian von Mammen
Patrick Lehner
Sven Tomforde

Keywords: ensambles; genetic programming; lawnmower problem; facades

Abstract:
The fast evolution of quadrotors paves the way for the application of robotic maintenance on walls and facades. In this work, we present an application-oriented modelling and simulation effort to evolve self-organising behaviour of a set of quadrotors. In particular, we rely on Genetic Programming (GP) to optimise control programs to collaboratively direct the route of autonomous quadrotors across a (vertical) rectangular surface relying on local knowledge only. In order to account for various real-world constraints, we made use of a three-dimensional, physical model that considers battery consumption, collisions, and rudimentary avionics. The evolved control programs that link sensory information to actuation could be executed on the Robot Operating System (ROS) available for numerous robotic systems. Our results show that very simple evolved control programs (moving left until the battery is drained, starting from a random location) perform better than those that we had previously engineered ourselves.

Pages: 16 to 21

Copyright: Copyright (c) IARIA, 2016

Publication date: March 20, 2016

Published in: conference

ISSN: 2308-4197

ISBN: 978-1-61208-462-6

Location: Rome, Italy

Dates: from March 20, 2016 to March 24, 2016