Abstract
The analysis of failure time data has grown in sophistication over the last fifty years with advances in theory and computation. Early methods, the mathematics of demographic Life Tables, have serious limitations when applied to subjects where the event of interest, e.g. failure time of a robot in space, does not occur and sample sizes are small. Recent advances in statistical theory have developed related techniques collectively called Survival Analysis, which overcome the main problems of older methods. In Survival Analysis there may be censored observations, where failure never occurs or observations are lost before failing. The median value of a set of observations is used as the primary statistic, not the mean or average statistic. Extensions of Survival Analysis allow for regression-like tools for multi-variate analysis. Further challenges will involve the mixing of real data from space and simulation data from Earth to improve estimation.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
