An integrated approach to the development of an intelligent prognostic health management system

The general aim of a Prognostic Health Management (PHM) system is to generate information to meet objectives such as improved mission reliability, improved safety and reduced maintenance costs. Information is derived from data (e.g. acquired by sensors) through the application of knowledge. Knowledge can be represented in different forms and can be implicit or explicit. The form of representation depends on the type of technology fulfilling some specific PHM function. PHM is complicated and successful systems will be constructed using technology from many suppliers. Careful management of technology integration is required. Proposed standards have been written to allow different technologies to communicate. But there is also a need to define formal frameworks so that technologies are integrated in a coherent manner. This paper presents some practical applications of PHM technologies that fulfil functions within different layers of the PHM process. These applications are then used to explain how these different functional technologies integrate through a common probabilistic framework.