Two key questions that many people ask are, “What evidence exists that implementing analytics and IIoT will improve efficiency, and what are the key areas to make an implementation successful?”
There are a growing number of studies and published reports that provide insights into these questions. One key finding is that good results come from minimizing data leakage at all stages of the IIoT implementation process beginning with the initial sensor points, continuing through storage in the automation infrastructure (historian) to analytics in the cloud. Bad, or missing, data at the start of the process can badly skew the final results. McKinsey and Company reported on this as far back as August of 2014 in their report “Digitizing Oil & Gas Production”. More recently, the well-publicized use of GE’s Predix IIoT analytics platform by Columbia Gas Group highlights this growing trend. A key element in the overall solution is Columbia’s deployment of Stratus’ always-on ftServer platforms in their compression stations, to guarantee availability of the local SCADA/HMI/Historian along with absolutely no data loss.
Availability has always been a major concern for most automation applications, but the McKinsey paper illustrates that the success of IIOT applications also depends on no data loss and data integrity from the source onwards.