Article Abstract

	Title:	Development of an intelligent robotic system for the automation of a meat-processing task
	Author:	S.A. Hurd, D.A. Carnegie, N.R. Brown, P.T. Gaynor
	Address:	DeviceWorks, Industrial Research Limited, PO Box 2225, Auckland, New Zealand ' Department of Physics and Electronic Engineering, University of Waikato, Private Bag 3105, Hamilton, New Zealand ' DeviceWorks, Industrial Research Limited, PO Box 2225, Auckland, New Zealand ' Department of Physics and Electronic Engineering, University of Waikato, Private Bag 3105, Hamilton, New Zealand
	Journal:	International Journal of Intelligent Systems Technologies and Applications 2005 - Vol. 1, No.1/2  pp. 32 - 48
	Abstract:	A robotic system and specifically designed end-effector tool has been developed for the automation of a meat processing task. This tool replicates the operation of a human butcher, while providing improved aseptic conditions and an increase in the reliability, repeatability and quality of the process. The robotic system incorporates a neural network-based fault detection scheme, which uses a learning vector quantisation network to classify axial load signals. A proposed online adaptive learning algorithm will provide continuous optimisation of the robot cut-path, leading to the automation of the path tuning process. This will allow the system to respond to seasonal carcass variations automatically and maintain a consistently high level of performance.
	Keywords:	fault detection; path optimisation; neural networks; learning vector quantisation; meat processing robot; line search algorithm; biomechatronics; biorobotics; aseptic conditions; reliability; repeatability; quality; butcher; adaptive learning; robot learning; seasonal carcass variations; robot path planning; sheep meat.
	DOI:	10.1504/IJISTA.2005.007306
		Purchase this Paper         Comment on the Paper