Neural networks in manufacturing: A survey

Abstract

Artificial Intelhgence (AI) has been claimed to yield revolutionary advances in manufacturing. AI technology is applicable to the entire range of manufacturing activities. As an AI technique, expert systems have been wildly used in manufacturing for over two decades. Recently, another AI technique, namely, neural networks, are gaining more and more visibility and have been successfully applied in manufacturing practices. While most of the survey papers about AI in manufacturing have been focused on expert systems application, fewer attentions have been paid to neural networks. However, expert systems are less effective in the ever-changing , complex, and open system environment of today's manufacturing systems. On the other hand, neural networks are able to learn, adapt to changes, and can mimic human thought processes with little human interventions. They could be of great help for the present Computer Integrated Manufacturing (CIM) and the future Intelligent Manufacturing Systems (IMS). Therefore, it is necessary to present the state-of-the-art of neural networks in manufacturing. The objective of this paper is to update information about neural networks in manufacturing, which will provide some guidelines and references 'for the research and implementation. In this paper, the basic concepts of neural networks are briefly introduced, Then, a survey of neural network applications in manufacturing is provided. Finally, the projection of future trends is given to help make decisions concerning neural networks implementation in manufacturing today and to aid in guiding research for tomorrow.

Pre-production results demonstrating multiple-system models for yield analysis

“…Neural networks have been widely used in manufacturing studies. A small sampling of the applications in manufacturing, in general is discussed by Page et al [11]; Udo [12]; Dagli [13] and Huang and Zhang [14]. May [15] and Mahajan et al [16] discuss semiconductor processing applications and in a paper on plasma modeling Rietman [17] reviews and contrasts many neural network models for plasma processing in semiconductor manufacturing.…”

An intelligent system for predicting HPDC process variables in interactive environment

“…Artificial neural networks, with their remarkable ability to derive meaning from complicated or imprecise data, can be used to extract pattern and detect trends that are too complex to be noticed by either humans or other computer techniques (Huang and Zhang, 1993). A trained neural network can be thought of as an "expert" in the category of information it has been given to analyze.…”

“…This expert can then be used to provide projections given new situations of interest and answer "what if" questions. Other advantages of ANN include-(i) Adaptive learning: an ability to learn how to do tasks based on the data given for training or initial experience, (ii) Self-organisation: an ANN can create its own organization or representation of the information it receives during the learning phase and (iii) Real time operation: ANN computations may be carried out in parallel (Huang and Zhang, 1993;Masters,1993).…”

Recurrent neural network approach for cyclic job shop scheduling problem

“…Several researchers [8][9][10] have applied neural networks on scheduling problems. Many of these works focused on the types of neural network to use either for production or manufacturing scheduling problems.…”