Despite their seemingly impressive claims, current products for Condition Monitoring, Diagnostic and Decision Support Systems (CMD&D) do not provide the reliable bottom line information that end users and operators need. Instead they confuse the issue with gigabytes of logged trends, complex cause-effect matrices, fault signatures etc. The term “Intelligent Health Control” here refers to the next generation of such systems which provide usable information on:
• the existence and severity of faults;
• how their severity will progress with utilization;
• how this progress can be influenced or controlled.
In this paper the fundamental shortcomings of current approaches are discussed prior to introducing the basics of Intelligent Health Control in terms of fault models and how they can be used to close the diagnostic, prognostic and intelligent control triangle.
The industry will unavoidably shift towards an “information centric” view from the currently predominant “data centric” view. Gigabytes of performance trends will no longer be relevant. Instead, reliable bottom line information will be required on how to minimize or control the costs associated with machinery health degradation or faults. In order to keep the discussion real, the current state of the art of enabling technologies are discussed, including:
• Open Information Buses;
• Adding real time data server functionality to the control system;
• Computational Steering, Human-in-the-Loop Optimization (or semi-automatic problem solving);
• Fault Models;
• Faster than real time simulation;
• Neural Nets.