Ricardo A. Téllez

Evolving the walking behaviour of a 12 DOF quadruped using a distributed neural architecture

This paper describes how a distributed neural architecture for the general control of robots has been applied for the generation of a walking behaviour in the Aibo robotic dog. The architecture described has been already demonstrated useful for the generation of more simple behaviours like standing or standing up. This paper describes specifically how it has been applied to the generation of a walking pattern in a quadruped with twelve degrees of freedom, in both simulator and real robot. The main target of this paper is to show that our distributed architecture can be applied to complex dynamic tasks like walking. Nevertheless, by showing this, we also show how a completely neural and distributed controller can be obtained for a robot as complex as Aibo on a task as complex as walking. This second result is by itself a new and interesting one since, to our extent, there are no other completely neural controllers for quadruped with so many DOF that allow the robot to walk. Bio-inspiration is used in three ways: first we use the concept of central pattern generators in animals to obtain the desired walking robot. Second we apply evolutionary processes to obtain the neural controllers. Third, we seek limitations in how real dogs do walk in order to apply them to our controller and limit the search space.

Evolving cooperation of simple agents for the control of an autonomous robot

Abstract A distributed and scalable architecture for the control of an autonomous robot is presented in this work. In our proposal a whole robotic agent is divided into sub-agents. Every sub-agent is co ded into a very simple neural network, and controls one sensor/actuator element of the robot. Sub-agents learn by evolution how to handle their sensor/actuator and how to cooperate with the rest of sub-agents. Emergence of behaviors happens when the co-evolution of several sub-agents embodied into the single robotic agent is produced. It will be demonstrated that the proposed distributed controller learns faster and better than a neuro-evolved central controller.

Using a cognitive architecture for general purpose service robot control

A humanoid service robot equipped with a set of simple action skills including navigating, grasping, recognising objects or people, among others, is considered in this paper. By using those skills the robot should complete a voice command expressed in natural language encoding a complex task (defined as the concatenation of a number of those basic skills). As a main feature, no traditional planner has been used to decide skills to be activated, as well as in which sequence. Instead, the SOAR cognitive architecture acts as the reasoner by selecting which action the robot should complete, addressing it towards the goal. Our proposal allows to include new goals for the robot just by adding new skills (without the need to encode new plans). The proposed architecture has been tested on a human-sized humanoid robot, REEM, acting as a general purpose service robot.

Evaluating the use of robots to enlarge AAL services1

We introduce robots as a tools to enhance Ambient Assisted Living (AAL) services. Robots are a unique opportunity to create new systems to cooperate in reaching better living conditions. Robots offer the possibility of richer interaction with humans, and can perform actions to actively change the environment. The current state-of-art includes skills in various areas, including advanced interaction (natural language, visual attention, object recognition, intention learning), navigation (map learning, obstacle avoidance), manipulation (grasping, use of tools), and cognitive architectures to handle highly unpredictable environments. From our experience in several robotics projects and principally in the RoboCup@Home competition, a new set of evaluation methods is proposed to assess the maturity of the required skills. Such comparison should ideally enable the abstraction from the particular robotic platform and concentrate on the easy comparison of skills. The validity of that low-level skills can be then scaled to more complex tasks, that are composed by several skills. Our conclusion is that effective evaluation methods can be designed with the objective of enabling robots to enlarge AAL services.

Highly modular architecture for the general control of autonomous robots

The implementation in a robot of the coordination between different sensors and actuators in order to achieve a task requires a high formulation and modelisation effort, specially when the number of sensors/actuators and degrees of freedom available in the robot is huge. This paper introduces a highly distributed architecture that is independent from the robot platform, capable of the generation of such a coordination in an automatic way by using evolutionary methods. The architecture is completely neural network based and it allows the control of the whole robot for, in principle, any type of task based on sensory-motor coordination. The article shows how the proposed architecture is capable of controlling an Aibo robot for the performance of three different difficult tasks (standing, standing up and walking) using exactly the same neural distribution. It is also expected that it will be directly scalable for higher levels of control and general design in evolutionary robotics.

Guest Editorial Sensorimotor Contingencies for Cognitive Robotics

The sensorimotor approach to cognition states, that the key to bring semantics to the world of a robot, requires making the robot learn the relation between the actions that the robot performs and the change it experiences in its sensed data because of those actions. Those relations are called sensorimotor contingencies (SMCs). This special issue presents a variety of recent developments in SMCs with a particular focus on cognitive robotics applications.

Embodying cognitive abilities: categorization

In previous woks we have introduced a distributed neural architecture for the generation of complex behaviors in evolutionary robotics. In this paper we show how this architecture is able to create its own categories about the sensed world of a robot by direct interaction of the body with the environment. The distributed elements of the architecture cooperate to express the categories on an inner world that is easily accessible from the outside. We conclude the paper with an explanation of how the inner world created by the robot can be used to gain some insight into the mind-body problem.

Webots Simulator 5.1.7 . Cyberbotics Ltd. (2006). Available in different versions at different prices; CHF 3600 (

Webotsk [1, 14] is a commercial program for the simulation and prototyping of mobile robots. It is developed and supported by Cyberbotics Ltd. [1], a leading company in simulation software founded in 1998 as a spinoff from the Micro-Computing and Interface Lab (LAMI) of the Swiss Federal Institute of Technology in Lausanne (EPFL). The Webots software allows the user to create 3D virtual worlds for the simulation of robots and their environments, including physical properties like mass distribution, friction coefficients, and inertia. Simulated robots can have different locomotion schemes, including wheeled, legged, swimming, or flying robots and can be equipped with various sensor and actuator devices: IR distance sensors, motor wheels, cameras, servos, touch sensors, grippers, emitters and receivers, and many more. Webots is provided with a number of models and transfer libraries for some commercial robots, including the Sony Aibo, LEGO Mindstorms, Khepera, Hemisson, and e-puck robots. Webots can be used for different types of research, such as mobile robot prototyping for the automotive industry, multi-agent research in collaborative mobile robot groups, or adaptive behavior research with genetic evolution, neural networks, and artificial intelligence techniques. It also generates impressive-looking result presentations by using a built-in module that obtains screenshots and movies from the simulations. Developed beginning in 1996 and available for the Windows, Linux, and Mac OS X operating systems, Webots has become a reference program used by over 300 universities and research centers worldwide. The first version (v0.4.alpha) was released in 1997 [13] and was updated yearly: Webots v1.0 (1998), v2.0 (1999), v3.0 (2000). Version 4.0.beta1 was released in February 2003 and included a working cross-compilation for the LEGO MINDSTORMS RCX. Released on January 2005, version 5.0.0 has been updated integrating a multi-platform binary server for Aibo ERS7/ERS210 and the Aibo ERS7 cross-compilation system [12]. The current version (as of June 29, 2006) is Webots 5.1.7. Since Webots was originally developed as a research tool, full upward compatibility is not always ensured between versions; however, continuous user support for technical questions concerning the simulator is offered by Cyberbotics, and an active webots-users mailing list [10] for discussion among Webots users proves to be a useful tool. Cyberbotics promotes a Webots based Roboka [7] robot programming contest, which is a competition held on the Internet where a free version of the mobile robot simulation software is used. This allows the contestants to program the robots in Java for a Judo competition between two humanoid robots, which are models [15] of the Sony QRIO robot.