The Energy Transition Is Also a Digital Transition
By Andy Lubershane and Shawn Cherian
Why utilities? The energy transition is also a digital transition
For over a decade, the tech industry has been famous for its unofficial motto: “Move fast and break things” (This was actually the official motto of Facebook, in the early days).
The utility industry also has an unofficial motto, in truth — a mandate — which it has maintained for over a century: “Safe, reliable, and affordable.” In other words: “Move slowly, and don’t break anything.”
It’s time for these historically disparate industries to cross-pollinate worldviews. We will not linger on the tech industry’s issues in this article; suffice to say that tech might benefit from breaking… fewer things. Instead, we want to focus on how utilities can “move faster” — how they can borrow the best of what tech has to offer in order to become more agile, creative enterprises.
Utilities need to “move faster” in order to confront three critical challenges:
1) Our global civilization has about three decades to dramatically cut greenhouse gas emissions, but a lot of the heavy lifting must be done in the next ten years for there to be any hope. Utilities are the cornerstones of that effort.
2) Utilities run the electrical grid on which we are already highly dependent, and on which we’re set to become even more dependent as we address Challenge #1. The safe, reliable, affordable functioning of that grid is already facing an escalating array of threats — in both physical and cyber form — and utilities need to be sufficiently agile to take on those threats as they arise.
3) As a century-old industry that never gets any down-time to turn the lights off and retool, utilities are burdened by an especially high level of technical debt.
That third challenge is where we at Energy Impact Partners see an opportunity. Utilities have not historically been the top choice for tech startups looking to break into a new market vertical (maybe something about that “move slowly” motto…). But the urgency of the first two challenges means that utilities are on the cusp of becoming exactly the right vertical for new tech adoption. Our company would not exist if this major industry wasn’t ready to move faster, and when they do move, our partners alone represent over $65 billion in annual spend in North America and Europe.
The past decade(s) of utility technology strategy
Utilities face many of the same decisions as other enterprises when it comes to technology adoption. At a high-level, these decisions fall along three axes:
1) Suite or best-in-class? A “suite” approach to technology adoption means procuring as many related product functions as possible from the same vendor (or building them all in-house), as part of a cohesive solution set. At the other end of the spectrum, enterprises can pursue the absolute best possible product for each discrete bit of functionality, and somehow piece those products together into an integrated solution.
2) Horizontal or vertical tools? Horizontal solutions are designed to be agnostic to industry-specific systems, workflows, and use cases; they ought to work just as well out of the box for a bank as for a utility. Vertical solutions, on the other hand, are narrowly tailored to addressing industry-specific problems.
3) Build or buy? This one’s obvious. But there is one important note: Typically, enterprises won’t build horizontal solutions in-house. After all, if you’re going to build something yourself, you might as well build it exactly to spec!
Historically, utilities have leaned towards a suite procurement approach. When your mandate is “safe, reliable, and affordable,” buying an integrated suite of products from established vendors like Oracle, SAP, and GE tends to be the presumptive choice. Large, cohesive technology suites have historically been easier to set up and maintain than assemblages of point solutions, and they’re often perceived as the choice of least operational risk.
Moreover, the utility regulatory construct also tends to favor the suite procurement approach. For the most part, utilities are able to earn a rate of return on capital investment, but not on O&M expenditures; hence, the utility business model inherently favors paying up-front for a comprehensive suite, rather than paying the salaries of large teams of developers to continuously tighten the bolts on pieced-together or home-brewed systems.
On the question of horizontal versus vertical solutions, utilities have tended to lean vertical. When it comes to operational technology (OT) for monitoring and controlling electric power equipment, that’s often been the only option. For example, tools for identifying outages in the electric distribution system generally need to be built for purpose.
Build or buy? Utilities aren’t manufacturers. They don’t produce their own power generation equipment or transmission towers. Instead, utilities have long understood their role to be smart buyers, integrators, and operators of this type of equipment. As digital systems have increasingly come to underly physical energy infrastructure, utilities have mostly stuck to their “smart buyer/integrator/operator” core competency.
Emerging opportunities for enterprise software
As utilities have begun to transition from “move slowly and don’t break anything” to “move quickly and… we still can’t afford to break anything,” they’re being pulled in new strategic directions. From our perch at the intersection of the utility and tech markets, we see three emerging changes in utility tech roadmaps that presage new opportunities for enterprise software vendors.
First, utilities are experimenting with building some of their own digital point solutions. The impetus for these experiments was the rapid evolution and diffusion of machine learning techniques — and, to be frank — the extraordinary hype around artificial intelligence. For a time in the middle of the last decade, high-priced consultants and think-pieces alike espoused the view that “data was the new oil,” and that all enterprises needed to “become AI companies” in order to thrive.
For a time, it seemed as though any enterprise — utilities included — could become “manufacturers” of their own best-in-class machine-learning-based solutions. Unlike physical equipment, which patently requires huge economies of scale to produce cost-effectively, machine learning seemed like the sort of thing almost any mid-sized enterprise could produce on its own: just hire a few data scientists and coders, and start sniffing around the enterprise for use cases.
We’ve since learned that there are limits to the “build it in-house” approach to ML. It turns out that there are actually economies of scale in building ML solutions — specifically, in training algorithms on enormous datasets, incorporating expert feedback, and of course, integrating ML-based solutions into a plethora of established systems and workflows. Oftentimes, for use cases that are common across an industry, third-party vendors can assemble the highest quality solutions and, in effect, spread the cost of developing those solutions across scores of customers. That said, there are always going to be edge cases in which enterprises encounter problems so singular and specific to their own operations that those problems are better solved through in-house development.
Even for in-house analytics, though, there’s often still a need for third-party solutions to support in-house data scientists and business analysts. For example, data science “workbenches” or “platforms” are enterprise tools that take much of the hard-core coding out of the process, thereby putting the latest and greatest ML algorithms in the hands of subject matter experts — who probably have the statistical chops to interpret those models, but not to build them from scratch.
These tools bring us to the second new direction in which utilities are expanding their digital footprint: infrastructure. As utilities rethink their digital strategies, many are finding that their infrastructure needs are not so different from most other mid-sized enterprises. Next-generation solutions in analytics, storage, networking, security, and other “horizontal” enterprise software categories that have worked well in other verticals can be powerful force-multipliers for energy providers with a less well-established digital heritage. That said, we also occasionally encounter more vertical-specific infrastructure solutions — for example, integration platforms purpose-built to connect existing OT systems and applications.
One category that will be especially vital for energy infrastructure is security. Threats to industrial controls systems, networks, endpoints, cloud infrastructure, and IoT devices all represent opportunities for the startup ecosystem to address growing threats to our critical energy delivery systems. EIP’s unique position within this ecosystem allows us to build the connective tissue between innovative security companies and our limited partners.
Third, and perhaps most importantly, utilities are increasingly moving away from the suite approach to procurement. As discussed above, there are clear benefits to single-source procurement. But when it comes to specific applications — especially those rooted in ML — utilities are finding that point-solution vendors are able to build superior solutions, faster, while competing on cost with entrenched incumbents. The aforementioned economies of scale in building ML-based products appear to be captured much more easily by highly-focused vendors than those trying to manage increasingly unwieldy and far-ranging suites. Also, investment in digital infrastructure ought to enable utilities to add new point solutions with fewer integration headaches.
Adding “Clean” and “Agile” to “Safe, Reliable, and Affordable”
“Safe, reliable, and affordable” has worked well for over a century (we certainly don’t recommend replacing that with “move fast and break things”). But we know that energy infrastructure providers will be increasingly required to add “clean” and “agile” to their mantras. As they do, we believe that EIP represents a unique platform for bridging the gap between those strategic mandates — balancing the risk of innovation with the rewards of efficiency, productivity, and cost. Bridging that gap isn’t theoretical for us; it means doing the legwork of bringing together the most promising innovators across the technology landscape with the key decision makers who are building the tech foundation for the utility of the future. We’ve proven this model, investing in best-in-class point solutions for our industry, and we believe the time is right to do the same for enterprise infrastructure.
