Process analytics software mines historical production data to reduce variability
Yokogawa's software combined with analytical services is designed to help manufacturers stabilize quality even as the variability of feedstocks increases.
Yokogawa’s process data analytics platform is designed to help process manufacturers detect changes in product quality at an early stage of production. It looks for changes by analyzing process data, facility status information, operation history, and other historized data. The software operating at the site works in combination with an analytical service element. The process data analytics platform runs on Microsoft Windows and examines production operations looking at temperature, pressure, flow rate, liquid level and other process data combined with facility operations and maintenance collected by the plant information management system in the distributed control system (DCS) or programmable logic controller (PLC). The analytical tools look for deviations from normal conditions, and when these are spotted it will trigger an alarm that conditions exist that could contribute to a decline in quality.
The examination goes beyond just process data and considers materials, methods, machines, and manpower aspects of production. It is designed for a variety of process industries such as oil and gas, petrochemical, chemical, pulp and paper, primary metals, pharmaceutical, and food. Discrete manufacturers can use the system for applications such as automotive, glass, rubber, and electronics. Yokogawa began offering process data analytical services in 2008 and has been refining the capabilities since. This new commercialized version of the system continues to use the Mahalanobis Taguchi (MT) method of pattern recognition that is employed in multivariate analysis applications.
– Edited by CFE Media. See more Control Engineering diagnostic and asset management products.