Adaptive Behavior, 4 (2)

This article considers the problem of spatial representation for animat navigation systems. It is proposed that the global navigation task, or "wayfinding", is best supported by multiple interacting subsystems, each of which builds its own partial representation of relevant world knowledge. Evidence from the study of animal navigation is reviewed to demonstrate that similar principles underlie the wayfinding behavior of animals, including humans. A simulated wayfinding systems is described that embodies and illustrates several of the themes identified with animat navigation. This system constructs a network of partial models of the quantitative spatial relations between groups of salient landmarks. Navigation tasks are solved by propagating egocentric view information through this network, using a simple but effective heuristic to arbitrate between multiple solutions.

Key Words

animat AI; spatial representation; navigation; multiple subsytems; quantitative models

Pages 125-172

Monitoring Strategies for Embedded Agents: Experiments and Analysis

By Marc S. Atkin, Paul R. Cohen

Abstract

Monitoring is an important activity for any embedded agent. To operate effectively, agents must gather information about their environment. The policy by which they do this is called a monitoring strategy. Our work has focused on classifying different types of monitoring strategies and understanding how strategies depend on features of the task and environment. We have discovered only a few general monitoring strategies, in particular periodic and interval reduction, and speculate that there are no more. The relative advantages and generality of each strategy will be discussed in detail. The wide applicability of interval reduction will be demonstrated both empirically and analytically. We conclude with a number of general laws that state when a strategy is most appropriate.

Key Words

Monitoring; genetic programming; embedded agents; planning

Pages 173-199

Discovering the Competitors

By Luc Steels

Abstract

This article reports on an experiment that tests whether a particular representation of robotic control processes is adequate for capturing significant variations in robot behavior. These variations can then be explored by a selectionist mechanism that generates and tests variations. An ecosystem modeled after a physical robotic ecosystem is introduced. The ecosystem contains a robot that occasionally has to recharge, as well as competitors that take away energy from the total system. The robot has to discover that its viability requires combating the competitors.

Key Words

autonomous robots; learning

Pages 201-210

Biological Adaptations and Evolutionary Epistemology

By James H. Fetzer

Review of Darwin Machines and the Nature of Knowledge, edited by Henry Plotkin. Cambridge, MA: Harvard University Press, 1994.

back to TOC, back to top

Adaptive Behavior, 4 (2)

Adaptive Behavior

Volume 4, Number 2

Fall 1995

Table of Contents

Tony J. Prescott

Spatial Representation for Navigation in Animats