As the in-line measurement is unavailable, and fixture faults, machine tool errors, and datum errors couple together to make error propagation much more complicate, the research in assembly is not directly applicable to machining processes. Therefore, Stream-of-Variation methodology should be extended for machining processes. Huang et al. (2000, 2003) have established a state space model to describe variation propagation in machining processes. And a process level diagnostics was proposed based on SOV model. Further work is needed in this area