When I talk with utility organizations about their data problems, the conversation usually starts in the same place: the systems are old. The technology has been in the ground for decades, it was never meant to last this long, and if the infrastructure were newer, everything would be easier. But is this really where the problem lies?
In fact, in a recent survey of utility organizations, only 22% pointed to legacy infrastructure as their primary challenge. The other 77% pointed to data quality and standardization. That gap has stuck with me because it means a lot of the industry is focusing on the wrong problem. The difficulty isn’t that the systems are old, but that utility data is structurally harder to work with than almost anything I see in other sectors.
In my last post, I made the case that the biggest obstacle to getting value from utility data usually isn’t the technology sitting on top of it, but the disconnected data underneath. Today, I want to get specific about why that data is so hard to connect in the first place. If you work in utility data, GIS, IT, or operations, this should help clarify why an approach that works cleanly in banking or retail tends to fall apart the moment it meets a utility.
More systems, more complexity
A typical utility runs its operations across a set of systems that each grew up independently, built to do one job well with no thought given to how they might one day share data with the others.
- GIS holds the spatial model of the network, i.e. where every asset physically sits.
- EAM (enterprise asset management) system tracks the condition, maintenance history, and lifecycle of those assets.
- SCADA sits closer to the equipment, monitoring and controlling it in the field in real time.
- OMS manages outages: detecting them, tracking restoration, coordinating crews.
- ERP handles the business side, including finance, procurement, work orders, materials.
- Document repositories hold the rest: PDFs, scanned drawings, permits, and inspection reports that live nowhere else.
Each of these systems is good at what it does. Each was designed and purchased to solve a specific operational problem, often years apart, often by different teams answering to different budgets. Sharing data was never a requirement.
In a typical utility I work with, the GIS team and the SCADA team have never really had a reason to sit in the same room, let alone agree on how an asset should be named. So the same transformer might be one record in GIS, another in EAM, and a third in SCADA, with no shared identifier tying them together. Multiply that across hundreds of thousands of assets, and you start to see the problem. When Saint John Energy set out to connect Cityworks and ArcGIS, the hard part was getting the systems to agree on what they were both looking at.
5 reasons integration is uniquely hard in utilities
Plenty of industries run a lot of systems. What makes utilities different is a stack of complications that rarely appear all at once anywhere else.
-
- The OT/IT divide. Operational technology (SCADA, sensors, field devices) and information technology (ERP, EAM, GIS) were built for different purposes. OT is governed by uptime, safety, and physical control; IT by data management, reporting, and business process. They use different standards, different security models, and different ideas of what one system talking to another is even allowed to do. They were never designed to speak to each other, and in many organizations, they’re owned by teams that barely overlap. Bridging that divide means crossing an organizational and technical boundary.
- Real-time versus batch data. SCADA produces a continuous stream of live operational readings (voltages, flows, states, alarms) updating by the second. EAM and ERP work in a different tempo: periodic, structured records updated when a work order closes or an inspection is logged. Bringing those two worlds means reconciling something that never stops moving with something that changes on a schedule. That takes transformation logic most utilities have never automated, so it happens manually, if at all. Devon Energy’s work to make their spatial and SCADA data more accessible is a good example of what it looks like to close that gap.
- Spatial data as a unique layer. In the same survey, 59% of respondents named GIS and spatial data integration as a top challenge. This is the piece most enterprise integration tools aren’t built for. Utilities can’t understand their data without understanding where it is. Every asset has a location, and location changes the meaning of everything connected to it: two identical assets aren’t interchangeable if one sits at a substation and the other feeds a hospital. Location is an important context that reshapes every other data point. Tools designed for non-spatial industries treat it as an afterthought, if they handle it at all.
- Round-the-clock operational constraints. Many sectors can take a system offline for a weekend to migrate or reconcile data. Utilities can’t. The network has to keep running. Every integration effort has to work around live operations, which slows the work down and introduces risk.
- Legacy dependencies with institutional knowledge baked in. Systems that have been in place for decades accumulate custom integrations, workarounds, and undocumented processes. Knowledge often lives in people’s heads rather than in any specification. Replace or integrate that system, and you risk losing understanding that was never written down. That fear keeps a lot of aging systems in place well past their natural life.
What the data tells us
Across the sector, 77% cite data quality and standardization as their core challenge, 59% point to GIS and spatial integration, and 66% report that the resulting fragmentation directly hurts productivity and quality of work.
Gartner has found that 63% of organizations are unsure whether their data management practices are ready for AI, a sign that data readiness is a broad struggle across every industry now investing in analytics and automation. The difference is that utilities carry all of those general challenges and then add the spatial layer and the OT/IT divide on top. Whatever the average enterprise is wrestling with, a utility is wrestling with more of it, plus complications that are structural to the infrastructure itself.
What connected data makes possible
When systems are connected and the underlying data quality is addressed, a set of capabilities opens up that fragmented data keeps out of reach. Predictive maintenance becomes possible when asset condition, spatial context, and live operational data finally sit together. Outage analysis and storm response get faster and more accurate when the OMS, GIS, and field data align, the way FortisBC’s work on real-time wildfire monitoring shows. Vegetation management becomes proactive instead of reactive. Regulatory reporting stops being a manual scramble and becomes something the data can largely produce on its own.
None of that requires ripping out the systems a utility already depends on. It requires a way to move, transform, and reconcile data between them without forcing everything into one monolith. That’s the role FME plays for utilities, handling the spatial, real-time, and cross-system complexity that makes utility integration hard. But the platform is a means to an end. The end is connected data and everything it makes possible.
The challenge: structural complexity, not old infrastructure
The data problem in utilities isn’t fundamentally about old infrastructure, and treating it that way leads to expensive projects that solve the wrong thing. It’s about the structural complexity of what utility data represents: spatial, operational, real-time, distributed across disparate systems. Understanding that difference is the first step toward fixing it.
In the next post, I’ll look at one specific, expensive symptom of all this fragmentation that hides in plain sight at nearly every utility: the hidden cost of PDF-based business processes.
