With the recent proliferation of computers, automatic sensor technology, communications networks and sophisticated software, quality control applications have fundamentally changed. The data used for quality control increasingly are multivariate and sampled individually in time at a high sampling rate. Hence the data will typically be cross-correlated and autocorrelated especially if sampled quickly relative to the dynamics of the system being monitored. Indeed the data stream used as input for process monitoring algorithms will often be high-dimensional discrete time vector time series. Unfortunately traditional approaches of statistical process control often fall short in delivering effective answers under such circumstances. In this paper we discuss some of the challenges this type of data bring forth in statistical quality control applications. We specifically discuss the robustness or lack of robustness of certain standard estimators.