To measure a product characteristic, sometimes the product must be destroyed. For example, when measuring breaking strength, the product is stressed until it breaks. Because the test is destructive, there is only one observation per product unit. In such a situation, you can model product reliability using the Destructive Degradation platform.

The platform models how a (typically) nonnegative response changes over time. Observations are assumed to be independent and measure the value of the response and the time at failure. A large and flexible collection of pre-defined models is provided. The models include location-scale and log-location-scale distributions whose location parameters are functions of time. The models allow explanatory variables and additional parameters.

For background on destructive degradation and reliability, see Escobar et al. (2003) and Meeker and Escobar (1998).