Tag is a children's game based on symmetrical pursuit and evasion. In the experiments described here, control programs for mobile agents (simulated vehicles) are evolved based on their skill at the game of tag. A player's fitness is determined by how well it performs when placed in competition with several opponents chosen randomly from the coevolving population of players. In the beginning, the quality of play is very poor. Then slightly better strategies begin to exploit the weaknesses of others. Through evolution, guided by competitive fitness, increasingly better strategies emerge over time. 1. Introduction Many of us remember playing the game of tag as children. Tag is played by two or more, one of whom is designated as it. The it player chases the others, who all try to escape. Tag is a simple contest of pursuit and evasion. These activities are common in the natural world, most predatorprey interactions involve pursuit and evasion. Tag also includes an aspect of role-reversal, b...

Robust behavioral control programs for a simulated 2d vehicle can be constructed by artificial evolution. Corridor following serves here as an example of a behavior to be obtained through evolution. A controller's fitness is judged by its ability to steer its vehicle along a collision free path through a simple corridor environment. The controller's inputs are noisy range sensors and its output is a noisy steering mechanism. Evolution determines the quantity and placement of sensors. Noise in fitness tests discourages brittle strategies and leads to the evolution of robust, noise-tolerant controllers. Genetic Programming is used to model evolution, the controllers are represented as deterministic computer programs.

The robot programs generated by Genetic Programming (GP) are found to be 'brittle', i.e. they fail to work when the environment is changed. Perturbation has been used to improve robustness. By introducing perturbation during the evolution of robot programs, the robustness of robot programs can be improved. This paper analyses the cause of the difference of robustness between robot programs using the case of robot navigation problems. The analysis is based on the notion of 'trace' of execution. The result of the analysis shows that the robustness of robot programs depends on the reuse of the 'experience' that a robot program acquired during evolution. To improve robustness, the size of the set of 'experience' should be increased and/or the amount of reusing the 'experience' shouM be increased.

Genetic Programming is used to create a reactive obstacle avoidance system for an autonomous mobile robot. The evolved programs take a black and white camera image as input and estimate the location of the lowest nonground pixel in a given column. Traditional computer vision operators such as Sobel gradient magnitude, median filters and the Moravec interest operator are combined arbitrarily. Five memory locations can also be read or written to. The first evolved program is now controlling the robot.

Traditionally the systematic study of animal behaviour using simulations requires the construction of a suitable mathematical model. The construction of such models in most cases requires advanced mathematical skills and exact knowledge of the studied animal's behaviour. Exact knowledge is rarely available. Usually it is available in the form of the observer's linguistic explanations and descriptions of the perceived behaviour. Mathematical models thus require a transition from the linguistic description to a mathematical formula that is seldom straightforward. The substantial increase of the processing power of personal computers has had as a result a notable progress in the field of fuzzy logic. In this paper we present a novel approach to the construction of artificial animals (animats) that is based on fuzzy logic. Our leading hypothesis is, that by omitting the transition from linguistic descriptions to mathematical formulas, ethologists would gain a tool for testing the existing or forming new hypotheses about `why' and `how' animals behave the way they do.

A significant problem in Genetic Programming consists in ensuring the robustness or generality of the evolved code (its ability to work correctly on never before seen data). We examine approaches attempted so far, and propose a different solution, based on multiple training sets (to discriminate between possible interpretations), augmentation and refinement, which improves on both task specification and distribution of fitness. We then present some preliminary results on its application to a fairly canonical GP problem (wall-following behavior) that has been previously shown to be very brittle. Finally, we discuss issues in using additional information about the generality of individuals.

This paper proposes a method to improve robustness of the robot programs generated by genetic programming. The main idea is to inject perturbation into the simulation during the evolution of the solutions. The resulting robot programs are more robust...

Genetic Programming was used to create the vision subsystem of a reactive obstacle avoidance system for an autonomous mobile robot. The representation of algorithms was specifically chosen to capture the spirit of existing, hand written vision algorithms. Traditional computer vision operators such as Sobel gradient magnitude, median filters and the Moravec interest operator were combined arbitrarily. Images from an office hallway were used as training data. The evolved programs took a black and white camera image as input and estimated the location of the lowest non-ground pixel in a given column. The computed estimates were then given to a handwritten obstacle avoidance algorithm and used to control the robot in real time. Evolved programs successfully navigated in unstructured hallways, performing on par with hand-crafted systems.

This paper describes an evolutionary way to acquire behaviors of a mobile robot for recognizing environments. We have proposed AEM (Action-based Environment Modeling) approach for a simple mobile robot to recognize environments. In AEM, a behaviorbased mobile robot acts in each environments and action sequences are obtained. The action sequences are transformed into vectors characterizing the environments, and the robot identifies the environments with similarity between the vectors. The suitable behaviors like wall-following for AEM have been designed by a human. However the design is very di#cult for him/her because the search space is huge and intuitive understanding is hard. Thus we propose the evolutionary design of such behaviors using genetic algorithm and make simulations in which a robot recognizes the environments with different structures. As results, we find out suitable behaviors are learned even for environments in which human hardly designs them, a...

Natural aggregation processes such as the familiar flocking of birds have been accurately modeled using a simple, decentralized controller. Variations on this “boid” controller typically involve three or more control laws, each with an associated control gain and sensor range. In this paper, the boid controller is fitted with an additional rule designed to produce aerodynamically-efficient formations, such as those exploited by migratory birds and hypothetical unmanned aerial vehicles. A simple genetic algorithm is then used to optimize the control parameters for minimum power consumption in a flock of simulated birds. This report focuses on the development and utility of the flocking simulator as a fitness function for the GA. Preliminary results indicate that average power consumption can be significantly reduced with the modified, optimized boid controller.