Catálogo de publicaciones - libros

The colonies of certain species of ants, for example , exhibit changes in behaviour as the colonies grow older, despite nearly all of the individual ants being replaced each year [1]. The behaviour of older colonies is more stable, and they are more likely to avoid intraspecific conflict [2]. Gordon hypothesised that the reason for this is that a 3-4 year old colony is in the steepest part of its growth curve, i.e. the 4000 workers of the 3 year-old colony are feeding 6000 larvae, and that the aggression of individual ants is based on colony level food requirements. This study aims to model this phenomenon using an individual-based simulation. The results from model are compared with field experiments taken over a period of years at the study site in New Mexico [3,4]. The model provides support to the biological hypothesis by showing that both colony age and aggression of individual ants have significant effects on foraging ranges.

There are examples of robotic systems in which autonomous mobile robots self-assemble into larger connected entities. However, existing systems display little or no autonomous control over the shape of the connected entity thus formed. We describe a novel distributed mechanism that allows autonomous mobile robots to self-assemble into pre-specified patterns. Global patterns are ‘grown’ using locally applicable rules and local visual perception only. In this study, we focus on the low-level navigation and directional self-assembly part of the pattern formation process. We analyse the precision of this mechanism on real robots.

Traditional eye-in-hand robotic systems are capable of performing versatile manipulation, but generally they can only observe a restricted workspace. As regards eye-to-hand configurations, tasks can be controlled within the field of view of the vision system with accuracy up to the pixel resolution of the vision system. In this paper, the robot workspace is further expanded by allowing cameras to be actively controlled by pan, tilt, and zoom motion. This configuration can be applied to a mobile robot equipped with a binocular vision system. The manipulator for visual servo control purpose can be either mounted on-board or in fixed configuration. To enable large and flexible workspace visual servoing with precision under such a configuration, active hand-eye coordination must be assured. The control strategy is successfully validated through convergence analysis and simulations.

The human body has a complex skeleton, giving a very high number of degrees of freedom, and is actuated by a large number of elastic elements – muscles and tendons. As a consequence, it has extremely challenging dynamics. Conventional humanoid robots use reduced degrees of freedom and traditional stiff actuators, and so fail to capture or exploit the important dynamic aspects of the human body. It has proved possible to build robots that mimic the human body – anthropomimetic or ‘musculo-skeletal’ robots – but the control of such robots will require very different methods from those used in existing humanoid robots. This paper reports the results of a preliminary investigation of the control problems using SIMNOS, a physics-based model of the anthropomimetic robot CRONOS. The transient and steady state effects of load changes on two simple feedforward methods for maintaining arm posture are assessed. The addition of a feedback controller reduces the steady state effects considerably, but still shows oscillatory transient effects. However, by combining this feedback controller with a velocity-limiting feedforward element, it proves possible to make smooth and reasonably accurate changes of posture under conditions of constant load.

Recent development in tethered airfoil i.e. kite technology allows the possibility of exploitation of wind energy at higher altitudes than achievable with traditional wind turbines, with greater efficiency and reduced costs. This study describes the use of evolutionary robotics techniques to build neurocontrollers that maximize energy recoverable from wind by kite control systems in simulation. From initially randomized starting conditions, neurocontrollers rapidly develop under evolutionary pressure to fly the kite in figure eight trajectories that have previously been shown to be an optimal path for power generation. Advantages of this approach are discussed and data is presented which demonstrates the robustness of trajectory control to environmental perturbation.

A new biologically inspired approach to a flapping wing controller which benefits from morphological computation and a Reflexive Pattern Generator (RPG) was tested using a simple physically simulated 3D flying robot. In order to tackle the difficulty of generating robust flapping flight and its manoeuvre, the robot employs simplified flexible “feathers” which are modelled as a series of subpanels attached to the wing skeleton using nonlinear angular springs. The neural controller receives sensory inputs from each feather to let them participate in pattern generation, the robot can also “feel” aerodynamic forces on its wings. From the synergy of flexible feathers and their sensory reflexes, the evolved robot exhibited flight manoeuvre using asymmetric wing movements as well as its tail, and rapidly adapted to external disturbances even in the absence of visual sensors. The reduced stiffness in flight control arising from the wing flexibility is discussed.

The paper presents a method to guide the self-organised development of behaviours of autonomous robots. In earlier publications we demonstrated how to use the homeokinesis principle and dynamical systems theory to obtain self-organised playful but goal-free behaviour. Now we extend this framework by reinforcement signals. We validate the mechanisms with two experiment with a spherical robot. The first experiment aims at fast motion, where the robot reaches on average about twice the speed of a not reinforcement robot. In the second experiment spinning motion is rewarded and we demonstrate that the robot successfully develops pirouettes and curved motion which only rarely occur among the natural behaviours of the robot.

We examine the practical problem of a mobile autonomous robot performing a long-duration survey task, during which it must recharge its batteries periodically. We present a scalable, online, heuristic method that allows the robot to recharge efficiently, thus maximizing its rate of work. The method is a direct application of the model, which seeks to explain the behaviour of animals solving related problems. Simulation results suggest that the method performs very well compared to optimal and naive heuristic approaches.

Real organisms live in a world full of uncertain situations and have evolved cognitive mechanisms to cope with problems based on actions and perceptions which are not always reliable. One aspect could be related with the following questions: could neural uncertainty be beneficial from an evolutionary robotics perspective? Is uncertainty a possible mechanism for obtaining more robust artificial systems? Using the minimal cognition approach, we show that moderate levels of uncertainty in the dynamics of continuous-time recurrent networks correlates positively with behavioral robustness of the system. This correlation is possible through internal neural changes depending on the uncertainty level. We also find that controllers evolved with moderate neural uncertainty remain robust to disruptions even when uncertainty is removed during tests, suggesting that uncertainty helps evolution find regions of higher robustness in parameter space.

In recent years simulation tools for agent-environment interactions have included increasingly complex and physically realistic conditions. These simulations pose challenges for researchers interested in evolutionary robotics because the computational expense of running multiple evaluations can be very high. Here, we address this issue by applying evolutionary techniques to a simplified simulation of a simulation itself. We show this approach to be successful when transferring controllers evolved for example visual tasks from a simplified simulation to a comparatively rich visual simulation.