A study of current logic design practices in the automotive manufacturing industry

This paper presents the results of an observational study of industrial logic designers at Lamb Technion. The purpose of this study was to determine the current methods of logic design used in industry. These observations can be used to evaluate recent academic proposals of methods for generating controllers for machines, as well as Suggesting new methods. By observing the logic designers it was determined that designing logic for machining systems is substantially different than writing computer code, both in the specification and in the people who will design and use the system. In addition, recent academic developments in Discrete Event Systems are difficult to apply to the problems of industrial logic control. Even methods which researchers design specifically for industrial logic design are difficult to apply, both due to the lack of support for the methods, and the unsuitability of the methods to solve problems in this domain. The primary observations of this study are the following. The observed Logic designers need to at least: determine acceptable machine behavior, foresee potential error conditions, predict user behavior, and design the logic needed for a machine. This is a greater range of responsibilities than expected. Logic for one machine is generally copied directly from a previous project. However, copying logic generally involves manually retyping everything due to incompatibilities in the development environments. Customers continue to ask for more features, and it is unlikely that this trend will stop. Some of these features are extremely difficult to implement using existing methods. Such features include: detailed part tracking, more sophisticated user interfaces, and greater diagnostic ability. (C) 2003 Elsevier Ltd. All rights reserved.