Dependability Benchmarking & Prediction: A Grand Challenge Technology Problem

We propose the grand challenge problem of dependability benchmarking and prediction for real-time mission-critical systems (RTMCSs). Evaluating dependability would quantify the degree of reliance that could justifiably be placed on a critical system, even in the face of partial failures or exceptional conditions. A comprehensive result would require an inter-disciplinary approach embracing the entire product lifecycle. While there are significant technical hurdles to both assessing the dependability of individual elements and combining resultant measures, a viable approach must be found to ensure that the computing systems our society is coming to depend upon will be reliable, available, safe, and secure. The participation of several communities, including the Real Time Computing community, is vital to success fully address this challenge.