Industrial application of condition monitoring and diagnosis prospects and pitfalls?

Summary form only given. There is a considerable gap between academic research and industrial practice in the area of fault detection, fault isolation and fault-tolerant control. Reasons for this are manifold and will be discussed based on concrete examples from industry. On one side there are increasing requirements on operational excellence in process plants. On the other side there is only a minimal workforce available for operation and maintenance. Automation of fault detection and isolation on one side and fault-tolerance on the other side should therefore be self-evident. They are, however, not. The talk will discuss reasons for this and show examples from industry where those obstacles have been overcome. As a consequence, some possible directions for academic research as well as success criteria will be pointed out.