Prognosis can be defined as the course of predicting a failure of equipment or a component in advance, whereas prognostication refers to the act of prediction. The three main branches of condition based maintenance are diagnosis, prognosis, and treatment-prognosis, however prognosis is admittedly the most difficult. Also, this area has been the least described in literature and the knowledge about it in a maintenance management context is still poorly systematized.

To this day, formal professional attention to prognosis, in the field of maintenance management and engineering in the everyday care of machinery, is often relegated to a secondary status although the availability of prognostic information can considerably improve (e.g. reduce costs, maximize uptime) the performance of machinery and maintenance processes. Ideally, assessment of a prognosis of remaining useful life should be deliberate and explicit. In order to support the maintenance crew in the achievement of this objective an increasing amount of prognostic information is available. Over the last decade, system integration has grown in popularity as it allows organizations to streamline business processes. It is necessary to integrate management data from CMMS (Computer Maintenance Management Systems) with CM (Condition Monitoring) systems and finally SCADA (Supervisory Control And Data Acquisition) and other control systems, widely used in production but with a seldom usage in asset diagnosis and prognosis.

The most obvious obstacle in the integration of these data is the disparate nature of the data types involved, moreover several attempts to remedy this problem have fizzled out. Although there have been many recent efforts to collect and maintain large repositories of these types of data, there have been relatively few studies to identify the ways these datasets could be related and linked for prognosis and maintenance decision making. After identifying what and how to predict incipient failures and developing a corresponding prognosis, maintenance engineers must consider how to communicate the prediction. In this activity once again, technicians' psychosocial attributes and values may influence how they discuss prognoses with asset managers. Regardless of whether prognostic assessments are subjective or objective, however, technicians should consider two major points.

Firstly, the maintenance crew should clarify in their own minds the link, if any, between their prognostic assessment and their consequent decision making. Secondly, they should consider the ways that they and their assets might benefit from explicitly discussing how the prognostic assessment is linked with diagnostics and preventive maintenance recommendations. These and other steps that maintenance engineers should take in incorporating prognostic information into their decision making are discussed in this paper. The objective is to give an overview of how the integration of disparate data sources, commonly available in industry, can be achieved for maintenance prognosis and optimal decision making.