Improving Robotic System Robustness via a Generalised Formal Artificial Neural System

Author

Howells, Gareth ; Sirlantzis, Konstantinos

Author_Institution

Dept. of Electron., Univ. of Kent, Canterbury

fYear

2008

fDate

6-8 Aug. 2008

Firstpage

23

Lastpage

28

Abstract

A major concern for robotic guidance systems is that a temporary or permanent failure of a given sensor within the system will erroneously trigger a potential system failure state. This paper introduces a generalised artificial neural system which is capable of addressing such problems by means of the inclusion of a weight value able to incorporate a distinct failure value. This will serve to significantly improve the performance and reliability of the guidance system.

Keywords

learning (artificial intelligence); neural nets; reliability; robot vision; generalised formal artificial neural system; learning algorithms; permanent failure; robotic guidance system reliability; robotic guidance system robustness; sensor; temporary failure; Artificial neural networks; Computer architecture; Formal specifications; Logic; Mathematical analysis; Neurons; Programming profession; Robot sensing systems; Robustness; Sensor systems; Generalised Artificial Neural Network; Robot robustness;

fLanguage

English

Publisher

ieee

Conference_Titel

Learning and Adaptive Behaviors for Robotic Systems, 2008. LAB-RS '08. ECSIS Symposium on

Conference_Location

Edinburgh

Print_ISBN

978-0-7695-3272-1

Type

conf

DOI

10.1109/LAB-RS.2008.12

Filename

4599422