Utilities are buying, implementing, and using data-science solutions in increasing numbers, with growing confidence, efficiency and urgency, as they face the challenges and embrace the opportunities presented by the moment at hand. This moment is one keenly felt by all utilities, a time when they must juggle multiple priorities – decarbonization, reliability, equity – all while facing downward cost pressures and increased public scrutiny.
If this sounds difficult, that’s because it is. But data, and the unlocking of its value via data science, is turning out to be the solution to rising to this challenge. Utilities are indeed having a moment, and with data science they are getting a taste for how they can meet and exceed it.
'It Takes Two'
Last year, E Source hosted its data science conference with executive leaders from electric utilities around the country. But we didn’t convene this group to sell anything – we listened as utility leaders discussed shared challenges and how each is tackling them using data-driven decision support.
The utility industry isn’t one to chase the flashiest new object. It’s an industry that is careful in its embrace of technology, valuing experience over bleeding-edge innovation. But that’s where data science is changing the conversation. Data science is no longer brand new or even particularly risky. Just about every utility has now stood up some version of an in-house analytics team, whether that’s a full-blown group of data analysts or an FTE dedicating some of their time to benchmarking how other utilities are using data to solve a particular problem or set of problems.
Now that everyone has “a line in the water,” the conversation is shifting to “how to catch more fish” with data science. One of the answers is that teaming with outside resources to accelerate and magnify the positive impacts of data science across use cases is now not only accepted, but preferred.
Think about it. The advantages of partnering with an expert data-science resource to get the most from your data investments are many. First is speed-to-value. Your data science partner likely has access to approaches, and hands-on experience implementing them, that have worked for your peers, knowledge that can streamline the use of data science and AI in your vegetation and storm management, undergrounding, and capital deployment programs and initiatives on the grid-infrastructure side, and demand-response participation, EV adoption prediction, and data-driven equity program design on the customer-engagement side.
Embedded in the speed-to-value benefit is access to best practices. Namely, there is no reason for you to make the same mistakes others have already learned from. An experienced data-science partner knows many of the pitfalls before you approach them, and that’s worth its weight in gold. You get to learn from their innovations while avoiding the pitfalls – working with a partner who has successfully mastered the process multiple times with your industry peers enables you to be a “fast follower.”
There are many other benefits to teaming with a skilled data-science provider, a firm ready to augment the talents of your internal team and respect the differences that make your utility unique while also tapping into the commonalities all utilities share in making data work faster and more efficiently in service of your decarbonization, reliability, equity and cost goals. But the good news is our industry has already realized this. We don’t need to sell the idea that great data science is necessary to rise to this unique moment, or that it takes more than an internal team to do this optimally.
Senior utility executives already know “it takes two,” because they are already doing it and loving the results.
The Great Convergence
So how long will it take before the incremental gains utilities are making today using data science snowball into something radically better, something that not only moves the needle, but changes the game altogether? Based on the groundbreaking work E Source and others have done with utilities across North America and the sentiments expressed by utility executives at our latest summit, I would say we are quickly approaching that tipping point, at least on the grid operations side, where data science becomes the invisible driver of every operating decision.
Let me explain.
Through a data-science-based approach called risk-spend efficiency, it is now possible for a utility to ask – and answer with certainty – the question: “Where can I get the most value from/biggest bang for my buck for every million dollars in my O&M and/or capital expenditure budget?”
Over the last five years, we have built the platform, applications, and know-how that enable utilities to do granular scenario planning on a precise digital replica of their grid to measure the financial, reliability, and service impacts of every dollar they spend to maintain it. (Note: we have developed a similar approach to optimizing customer engagement, one fundamentally different than anything else in the industry, called “Audience of One,” but that will have to wait for another article).
At its heart, this approach builds on something we have learned over time: the data and algorithms needed to “solve” a vegetation management optimization question are very similar to those needed to optimally place avian guards or underground wires across the grid. This commonality has led some of our most successful clients to move quickly from standing up one data-science project to two, then four, then 16, building game-changing momentum – and results that impact service levels, reliability, and the bottom line – with every success. Once the foundation of the digital replica is in place, the variety of algorithms needed to optimize activity on that replica – then on your physical grid – are either readily available, readily configurable, or ready to be developed quickly based on the need.
I call this rapid and exponential increase in the use and usefulness of data science “the Great Convergence” and see its impending approach – if not every nuance of its application and impact – as inevitable and transformative as the Industrial and Mobile Revolutions of the last two centuries.
That may seem like a big claim, but the utility industry is at its heart an industry of fast followers. Once something works, resistance to it falls. Nearly every utility now has at least one data-science line in the water. Many are catching fish. Some are now catching fish by the boatful.
It won’t be long before this bounty, the proven benefits of data science, is part of the narrative uniting all utilities in their quest to becoming more environmentally responsible, reliable, and equitable on their journey to becoming the Sustainable Utility. Data science and, by extension, AI are here today for utilities, and there isn’t a downside to the improvements they will drive.Tom Martin leads the Data Science team at E Source and is a key partner to utility clients as they look to harness data science and AI to solve critical business challenges. Prior to E Source, Tom served as Managing Director of Data Science at TROVE; earlier he led the implementation of new technology and analytics at PG&E to reduce the company’s operational costs, improve safety, and increase grid reliability. He can be contacted at [email protected]