A Fuzzy Intelligent System for End-of-Line Test

Yi Lu Murphey, Tie-Qi Chen, Jianxin Zhang, Jacob Crossman, and Brennan Hamilton
Department of Electrical and Computer Engineering
The University of Michigan-Dearborn

Abstract

     The success of the U.S. motor vehicle industry very much depends on the quality of the products it produces. As automotive electronic control systems have become more advanced and sophisticated in recent years, malfunction phenomena have also become increasingly more complicated. It is well recognized in the automotive industry that effective vehicle diagnostic systems will play a key role in the competitive market of the new century. In order to meet this challenge of improved quality control and diagnostics, the major US automotive companies are in the process of launching end-of-line test systems at every North American assembly plant.  Part of the end-of-line test system is designed to collect and analyze Electronic Engine Controller (EEC) data while the vehicle is dynamically tested.  Operators drive the vehicle through a preset profile and the vehicle is either passed or failed according to the data collected during the tests. The pass/fail decision is made based on two information sources – an EEC on-board tests and an EEC off-board test that is performed by the vehicle test system on EEC generated data.  Our Fuzzy Intelligent System is focused on automating the off-board testing process to obtain faster and more reliable test results than are currently realized by line engineers.