Expert Systems, A Decade of Use for Used-Oil Data Interpretation

摘要

Pressed by the need to improve oil analysis performance, some equipment operatorshave increased sampling frequency (shortened intervals) in order to increase the probability of early fault detection. As a consequence, laboratory labor costs increased considerably quadrupled in some cases. Over the past 10 years, expert systems have been increasingly used to compensate for the increased processing time by automatically interpreting sample data in near real time, improving evaluation reliability and minimizing the associated labor costs. A properly designed expert system can quickly review all recorded equipment and sample data, while keeping the analysis time and costs within acceptable levels. These systems greatly increase data interpretation consistency, and can generate significant returns-on-investment. This paper presents an overview of several of these systems and the general principles utilized in their development.