The PI System, Raw Data and Advanced Analytics: A Metaphor to Explain It All

At OSIsoft, I essentially serve as the CMO, or chief metaphor officer. I come up with analogies to make it easier for people--and particularly people not familiar with our technology or operational technology in general-- to understand the PI System and how to use it. (A die hard engineer might think, “Metaphors? I could teach a rhesus monkey to create those for half the price.” Actually, their paws aren't strong enough to depress the space bar.)

One of the big questions we're asked is where does the PI System end and third-party analytics begin. Can I use the PI System to analyze and reduce my energy consumption? (Yes, Alcoa and others have.) Can I use it to compare the performance of 70 plants? (You could, but it makes more sense to funnel all that data into OCS and enlist analytics.) The PI System can be used to serve up data predicting some of the primary wind turbine failures a week in advance (Sempra Renewables) or in combination with SparkCognition predict a wider range of problems off a more confusing range symptoms 60 days in advance (Invenergy.)

So here's the metaphor:

Imagine you're building a house. Your sensors, SCADA systems, IoT gateways and other devices are the equivalent of building supply vendors. They produce pipes, conduits, drywall, bricks, wires, faucets and all of the other parts that go into building a structure similar to how an IoT gateway might deliver ambient temperature readings updated every 15 minutes while B&K Vibro might capture vibration signals ever 15 microseconds to look for early symptoms of pending failure.

Building supply makers, however, don't assemble them: drywall has to be cut and the ceiling fans and ducts better be in place beforehand or there will be a change order. Thus, raw data is similar to having Home Depot Pro delivery truck vomit a random and overwhelming collection of materials on your driveway. You could try to assemble them quickly on your own, but the results might look like:

sea hut

The PI System is the contractor. It fits everything together and creates the house. The raw building materials haven't changed. They've just been sawed and organized in a way that makes the whole more valuable and useful than the parts.

The organization makes it easy for the occupants to solve problems on their own. The thermostats read 72 everywhere except the kitchen, where it's 97. That spike could mean a problem with the HVAC system or a fire. Either way, people know to check out the problem. The sound of running water? Time to jiggle the faucet and if the persists--and the higher costs show up on your bill-check everything.

But what about converting the third bedroom into an office? Would it add to the value of the house? Does it make more sense to add onto the back of the house to get more square footage or eliminate the hallway closet near the front door that people seem to only use for old coats?. Now you're in the realm of analytics. The PI System can help you lower energy, but someone like Canvass Analytics can map out what will happen if you go with different types of insulation.

And when you put it all together your life looks more like:

house

Powered by Translations.com GlobalLink OneLink Software