Location : SYMBRION REPLICATOR >
Print

Open Source


Robot3D simulator.


The simulator is tailored to physical simulation of multiple robots. The robots can be equiped with sensors (e.g. camera). Currently, all the hardware
platforms used in S/R projects are implemented: KIT robot, SSSA robot and Active Wheel. The project is hosted at Launchpad.


LibCCD.


This is new collision detection library for purpose in physical simulation. It allows to detect collisions between arbitrary pairs of convex shapes.
Beside that, the penetration vector between colliding objects can be computed, thus the library can be used in physical simulation. The library can be downloaded
here. The library was integrated into ODE physical engine. If you want to use it within ODE, compile ODE like this:

svn co https://opende.svn.sourceforge.net/svnroot/opende/trunk
cd trunk
./autogen.sh
./configure --help

Now you can see several options for collision detection.
--enable-libccd
  • This turns on libccd collision detection system, except box-cylinder collidor
--with-box-cylinder=libccd
  • Adds box-cylinder collider from LibCCD

./configure --with-box-cylinder=libccd
make
make install



Open Science


Appearance of multicellular structures is one of the greatest moments in the history of life. The rise of multicellular from unicellular is a huge evolutionary step, however we do not exactly know how multicellular organisms appear and which mechanisms take part in this phenomenon. We know multicellular organisms are self-adaptive, self-regulative and self-developing, however we do not know its evolutionary origin and developmental organization.

Robotics is seen by many researchers only as a tool to improve productivity. However, robotics can also be used as an instrument to explore unknown and unclear issues in nature, to model elements of organic life, to experiment with self-development and evolution, and to propose plausible explanations. The great vision, which consolidates many interdisciplinary researchers, is a vision of self-adaptive, self-regulative and self-developing robots that reflect multicellularity in nature — a vision of artificial robot organisms. Like multicellular beings, these artificial organisms consist of many small cell-modules, which can act as one structure and can exchange information and energy within this structure. Moreover, these structures can repair themselves and undergo evolutionary development from simple to complex organisms.

The vision of artificial multicellularity is very challenging. For design and production of mechatronic artificial cells, experts in robotics are required. An artificial organism, seen technically, is a large distributed computational system - it requires software engineers to program a basic functionality as well as to develop computer simulations. Organisms can autonomously harvest environmental energy and distribute it between cells - this is a deal for specialists in energetic homeostasis. Each of these cell-modules possesses dozens of different sensors. When many modules get integrated into one system, the problem of sensor fusion, distributed sensor processing, world modeling and developing a cognitive functionality become a reality. Adaptation requires aspects in machine learning and in adaptive control. Following a genetic bio-inspiration, artificial organisms possess a genome - this involves biologists, specialized in genetics. Artificial organisms should protect themselves - it requires an artificial immune system. To handle genetic issues, there is a demand on specialists in artificial evolution - computer scientists, as well as in natural evolution - ecologists and reproduction biologists. Finally, collectively working robots create social structures and are capable for evolving pre-semantic languages - this requires researchers of artificial cultures.

This section is devoted to further "open-end" discussions about such issues like artificial evolution, artificial genetics, morphogenese and morphodynamics, self-organization and others.

Artificial Immune System


  1. Overview of Immune System Biology
  2. Artificial Immune System
  3. Examples Scenario Dealt by AIS
  4. Online adaptation to compensate for erroneous behaviour
  5. Proposed Implementation of the AIS Algorithm
  6. Pseudocode for the Artificial Lymph Node Architecture

Artificial Evolution


This section is currently open only to registered users.

  1. SYMBRION EVOLUTION
  2. Artificial Evolution
  3. Genetic-based mechanisms group
  4. Genetic Draft
  5. Open discussion about artificial evolution


Learning


This section is currently open only to registered users.

  1. Machine Learning functionalities
  2. LogSystem
  3. Preparing Prague Meeting



Created by admin. Last Modification: Thursday 09 of June, 2011 10:23:27 CEST by vojta.

Menu

Home