Capturing Operational Value Through Data Analytics

Data analytics has undergone tremendous development in the last few years with many new techniques and algorithms. New applications are announced almost daily. However relatively few of these announcements have been in the process industries. Part of the issue is the characteristics of process plant data. It contains a wide mixture of time stamped data types
â?? numeric process variables values, commercial transactions, textual information, geographic equipment location records, frequency spectrums from special laboratory equipment and rotating equipment vibration measurements, and video/ audio recordings. However, the data is natively of poor quality and is not necessarily well structured. Process instrument readings drift
and noise corrupts the measurements. Even when the actual measurements are good, the statistical properties are not â?? i.e. process data is usually non-stationary, serially auto-correlated and cross-correlated. Another factor is that the potential economic payout is sometimes difficult
to estimate in advance leading to a lack of funding for the investments required. In this presentation potential application areas that would likely have high paybacks are identified along with the techniques required to attack these areas through plant data analytics. Significant additional value and margin can be gained by improving the accuracy of forecasts of future plant behavior, including potential production and supply chain alternatives, early detection of potential equipment problems, and product quality issues. In this presentation,
actual case studies will be used to illustrate the impact of these new tools on plant productivity
and margins.