Over the last decades the reliability, availability and safety of plants has become increasingly important for a number of reasons. One of the reasons is the increasing size and complexity of the plants, both onshore and offshore. The cost and safety risks have increased with the size, and the complexity has led to extensive automation. The human operator is not able to control the entire plant, nor able to assess its running condition, without the aid of this automation. However, the amount of data available to the operators is enormous. Many parameters, variables and alarms are available simultaneously for evaluation. Hence the need for automated evaluation of the data, or enhanced condition monitoring (sometimes incorrectly referred to as 'intelligent' condition monitoring). In 1992 the ICMOS (Intelligent Condition Monitoring Systems) research project was well underway for diesel engines used for propulsion and power generation onboard ships. Trying to extend this research and find other application for the results and experience led to research on condition monitoring for compression refrigeration plant, as these plants are often used and also often not well understood by the marine engineer. The initial, and necessarily very general, research assignment for this project was to generate knowledge required for the condition monitoring and fault diagnosis of compression refrigeration systems on board ships.