The University of Edinburgh -
Division of Informatics
Forrest Hill & 80 South Bridge

Research Paper #691

Title:A Dynamic Model for Autonomous Vehicle Navigation
Authors:Kick,A; Walker,AV; Fisher,RB
Date:Apr 1994
Presented:Submitted to SAB'94
Abstract:In recent years, much progress has been made in robotics research by employing ideas taken from biology. In particular, inspiration has been drawn from ethologists' study of insect navigatory capabilities. The long distance foraging behavior of honey bees (Apil mellifera) has been well studied and models of bee navigation have been suggested and simulated. The most popular of these models (due to Cartwright and Collett), proposes that honey bees navigate via a static image comparison process wherein motion vectors towards a target are calculated based on the differences between an image currently projected on a bee's retina and a remembered snapshot - taken from a similar place in the environment. The current work describes our implementation of such a static localization model on a mobile robot and discusses our conclusion that a dynamic model of bee navigation, based on optical flow, is more plausible. We propose a dynamic model based on comparison of apparent speeds of edges.

[Search These Pages] [DAI Home Page] [Comment]