Mechanical Systems and Signal Processing (2002) 16(4), 487–546 doi:10.1006/mssp.2001.1460, available online at http://www.idealibrary.com on REVIEW ON-LINE AND INDIRECT TOOL WEAR MONITORING IN TURNING WITH ARTIFICIAL NEURAL NETWORKS: A REVIEW OF MORE THAN A DECADE OF RESEARCH Bernhard Sick University of Passau, Chair of Computer Architectures (Prof. Dr.-Ing. W. Grass), Innstr. 33, 94032 Passau, Germany. (Received 18 December 2000, accepted 5 September 2001) The supervision of tool wear is the most difficult task in the context of tool condition monitoring for metal-cutting processes. Based on a continuous acquisition of signals with multi-sensor systems it is possible to estimate or to classify certain wear parameters by means of neural networks. However, despite of more than a decade of intensive scientific research, the development of tool wear monitoring systems is an on-going attempt. This article aims to investigate, why it has not been possible to develop appropriate monitoring systems up to now. In order to describe the ‘state of the art’, 138 publications dealing with on-line and indirect tool wear monitoring in turning by means of artificial neural networks are evaluated. The article compares the methods applied in these publications as well as the methodologies used to select certain methods, to carry out simulation experiments, to evaluate and to present results, etc. As a conclusion, possible directions for future research in this area are pointed out. Many of the recommendations are valid for other machining processes using tools with or without defined cutting edges, too. # 2002 Elsevier Science Ltd. All rights reserved. 1. INTRODUCTION Manufacturing processes like drilling, milling, or turning can be optimised significantly using reliable and flexible tool monitoring systems. The most important tasks in this context are [77]: * the fast detection of collisions, i.e. unintended contacts between the tool or the toolholder and the workpiece or components of the machine tool, * the identification of tool fracture (breakage), e.g. outbreaks at cutting edges, and * the estimation or classification of tool wear caused by abrasion or other influences. While collision and tool fracture are sudden and mostly unexpected events which require reactions in real-time, the development of wear is more or less slowly proceeding. This article focusses on the determination of wear, the most difficult of the three tasks. The importance of tool wear monitoring is implied by the possible economic advantages. By exchanging worn tools in time (considering the current machining process such as rough or finish turning, for instance), it is possible to avoid the production of waste. Furthermore, tools costs can be reduced noticeably with a precise exploitation of a tool’s lifetime. With an accurate estimation of tool wear it is even possible to adjust the tool position in order to meet geometric specifications and to control the tool wear rate in order to guarantee a certain surface quality of the workpiece (roughness). Ghasempoor et al. 0888–3270/02/ þ $35.00/0 # 2002 Elsevier Science Ltd. All rights reserved.