See this Comm Center thread by Craig Young. He wants a summary of the number of failed images at the end of a mass plate solve Solve.txt log.