We—meaning both the world in general and us here at Oracle specifically—have teams of data scientists who essentially play “what if?” all day long.
You remember that “what if” game you played as a kid: What if I could fly like a bird and take my dad to work every day so he didn’t have to drive through all that traffic? What if dogs could talk and I could send my dog next door to see what the neighbors are having for dinner and prove to mom that not every family is having liver? What if I had a robot hand and could fly and could pick up our house and move it to Disney World so I could visit every single day forever?
Our data scientists play “what if” the adult way, layering on planning and logical and strategy: They’re building product models and working with other data scientists to help them build their own models. They’re using modern open-source libraries, sorting petabytes of data and working with utilities to collect ground truth—the data that helps machine learning models truly learn and improve as they go along.
It’s all serious stuff, but, let’s be honest: It’s also fun. “What if” thinking is tons of fun, and if that “what if” thinking can become reality—well, that’s fun with a side order of productivity. And that’s what every one of us is aiming for with our “what if” questions.
Here are some “what if” questions we’re mulling right now.
What if you could actually predict, in advance, the damage that a forecasted storm will have on your network? You could then proactively move equipment and work crews into place based upon where the damage is predicted to occur, and be able to immediately begin repairs as soon as it is safe to do so.
What about smart home thermostats and peak load events? What if we could enable automated in-home device control based on utility price and event triggers? Here’s how that works: When your utility runs a peak pricing event day, you tell your customers to reduce energy consumption from, say, 2 p.m. to 5 p.m. in exchange for a rebate on their bill. That pricing trigger can automatically update the customer’s thermostat—raising the temperature at which the air conditioning kicks in by 4 degrees, for example—or automatically turn off the lights, or turn off major appliances like pool pumps or electric vehicle charging.
The traditional, linear model of generation/transmission/distribution/consumption is rapidly being replaced with a circular system allowing for multi-directional energy flow, within which customers can actively react to changing system demands and provide valuable services back to the grid.
If you think about it, it’s a brilliant example of “what if” thinking. As solar panels have decreased in price, customers are investing in them, wanting to be more involved with their energy consumption and even in the type of energy they are consuming. Not too long ago, there were only two options available to those customers: living completely off the grid, which was an enormous investment in battery technology, as well, or net metering. But what if technology could provide a way for those customers to be active participants of the grid?
The customer-centric, digital energy grid does precisely that: It’s a new model of the distribution grid from generation to transmission to distribution all the way out to consumers’ distributed energy resources at the grid edge. The “grid platform” leverages this digital energy model to optimize both the controllable distribution grid AND distributed energy resources to provide a safe, reliable and resilient distribution grid.
Here, I want to give you one final example of what happens when you take an outside of-the-box approach to connecting specific applications across the grid platform. So, what if you could tie your network management system and your demand-side management program application together with an IoT control system to identify load capacity issues and at-risk transformers and feeders, and then target the customers connected with those devices with demand response programs?
Here’s one way in which this could play out: Say a large penetration of rooftop solar is predicted to cause intermittency on a partly cloudy day, causing voltage violations in a specific neighborhood. Once this network constraint is identified, the network management system identifies available DER resources that can mitigate the constraint. The utility’s demand side management application then engages those customers to participate in the event, and requests permission to control their DER device for the event. (This can be done a week in advance of the actual event, based on “look ahead” forecasts of future grid constraints.) The network management system then sends the control signals to the optimal customer locations and, if additional response is then needed, sends out control signals to an additional set of opted-in customers. Following the event, the utility’s demand side management tools would provide settlements to reward the participants who did what they agreed to or penalize those who may have failed to meet their agreement.
This is the utility distribution grid of the future—one that allows the customer to be an active participant, and optimizes that participation, managing distributed generation, EV charging, grid resiliency and restoration after outage events and is agile enough to deliver innovative new products and services in the future.
In the world of technology, we’ve found, taking this unique thinking approach has shortened the innovation journey from “what if” to “what next.”
This is part two of a two-part series on innovation and planning. Read part one here.
Ready to get more in-depth on outside-of-the-box thinking and innovation? Dive into our recent innovation blueprint we created in partnership with Navigant. You can get that for free right here.