Animal behavior as a paradigm for developing robot autonomy

Tracy L. Anderson, Max Donath

Research output: Contribution to journalArticlepeer-review

64 Scopus citations

Abstract

We have been examining naturally occurring examples of autonomous systems in order to identify characteristics that might provide insight into our research on autonomy. In the first part of this paper, we review relevant research which has occurred in the area of animal behavior. Based on certain observations, we have proposed a number of primitive reflexive behaviors which are then used to develop several useful emergent behaviors. These emergent behaviors were demonstrated on a simulated mobile robot and then successfully implemented on Scarecrow, an actual robot. Scarecrow allows us to demonstrate that behavioral control strategies do indeed provide us with a powerful strategy for robust operation in dynamically changing unstructured environments in which one cannot impose unrealistic expectations on the performance of the machine or its sensors. A consequence of this is, that given the unpredictability of human actions, such behavioral control strategies may facilitate the safe interaction of man and machine.

Original languageEnglish (US)
Pages (from-to)145-168
Number of pages24
JournalRobotics and Autonomous Systems
Volume6
Issue number1-2
DOIs
StatePublished - Jun 1990

Bibliographical note

Funding Information:
This research has been supported with funds provided in part by the University of Minnesota Productivity Center, the CIM Consortium, the CyberOptics Corp., and the Office of Naval Research under Contract N0004-85-C-0847. We also appreciate the loan of equipment by MTS Systems Corp. and CyberOptics Corp. We would like to thank Jay Talbott for his efforts in desiL~-ningl arge portions of Scarecrow and for orchestrating the experiments.

Keywords

  • Emergent reflexive behaviors
  • Ethology
  • Mobile robots
  • Robot control
  • Stimulus response characteristics

Fingerprint

Dive into the research topics of 'Animal behavior as a paradigm for developing robot autonomy'. Together they form a unique fingerprint.

Cite this