R-DAT's non-homogeneous module provides an advanced analysis capability that allows users to construct generic prior distributions for their failure rate and failure probabilities.

In these non-homogeneous analysis, in risk analysis applications also known as population variability analysis or the first stage in two-stage Bayesian analyses, generic priors are constructed by analyzing the component-to-component or application-to-application variability of failure rates and probabilities. The variability is assessed based on information in the form of

• Field and test data: run-time and demand information resulting from actual operation or testing.
  
  
• Engineering judgments: estimates, along with uncertainty bounds, provided by expert engineers.

R-DAT is unique in that it can analyze any mix of data and estimates available to the user. The efficiency of R-DAT's algorithms have furthermore made it possible to replace discrete mesh priors, in use in other non-homogeneous analysis codes, by continuous priors. This eliminates inaccuracies due to approximations introduced by the discretization. Lognormal, Beta, and Gamma variability models are supported.

The advanced analysis capabilities of the non-homogeneous analysis module are controlled using an easy-to-use graphical user interface, that reduces the operation of setting up and executing the analysis to a series of simple mouse-button clicks. Results are made available in the form of charts and tables, which can easily be transferred to other environments such as Microsoft Word and Excel.