Everyone has a metaphor for data analytics that’s supposed to explain and clarify it, but they really don’t and mostly fall short. Its essence is simply turning raw, unhelpful information into useful guidance, efficiencies and value.
Many software packages and digitalized solutions promise to help, with a boost from apparently countless artificial intelligence (AI) tchotchkes. These are just the latest time-savings steps for moving process analytics away from clipboards, manual data entry and archiving, through paper charts recorders. Data acquisition (DAQ) hardware and spreadsheets, and onward to today’s computers, software-based historians, and Internet- and Ethernet-enabled, cloud-computing services.
However, just as the most precise instruments and controls can’t make up for badly installed or broken sensors, uncalibrated instruments, out-of-tune signals and other blurry inputs, data analytics can’t be effective if users don’t pick and apply appropriate tools guided by specific problems they want to solve or a practical goal they want to achieve.
- Tying together analytics locations, formats and functions: Hargrove Controls & Automation shows how users’ needs should determine analytics types and deployment.
