Executive Summary

Electric utilities across the U.S. and Canada are turning to undergrounding of distribution networks at unprecedented scale — driven by extreme-weather resilience, wildfire mitigation, and aging infrastructure. ~20% of U.S. distribution line miles are already underground, with major IOUs (PG&E, FPL, Dominion, WEC, Eversource) running multi-billion-dollar undergrounding programs targeting 50% underground by 2040. But undergrounding is only half the story. Out of sight, out of mind — utilities have less granular knowledge of underground asset condition than overhead, and yet underground failures cost 3–4× more than overhead and outages take longer to locate and restore.

Table of Contents

1 · Trend Toward Undergrounding — Utility Initiatives in US and Canada

Undergrounding Momentum in the US. Roughly 20% of America's distribution lines are underground today, up from ~18% in 2009². Many utilities have historically built new residential developments with underground utilities, and now several large utilities are embarking on aggressive undergrounding programs. For example:

• Pacific Gas & Electric (PG&E) – In wildfire-prone California, PG&E announced a $15–30 billion program to underground 10,000 miles of distribution lines (about 12% of its system)⁶⁷. By 2026 they plan ~3,600 miles underground, one-third of the goal⁶. This is part of a wildfire mitigation strategy to stop ignitions, reduce outages, and improve resiliency⁸. PG&E's CEO has described rebuilding the system "from the underground up."⁹
• Florida Power & Light (FPL) – FPL is ramping to ~$1 billion per year by 2025 to convert overhead laterals to underground¹⁰. Their Storm Secure Underground Program (SSUP) proved so successful in improving storm resiliency that Florida's regulator has allowed other utilities (Duke Energy Florida, Tampa Electric) to launch similar programs¹⁰. FPL justified SSUP by reduced restoration costs (especially avoiding major hurricane damage) and lower ongoing O&M¹¹. During Hurricane Irma (2017), underground lines had far fewer outages (mostly caused by debris on overhead lines)¹².
• WEC Energy Group – In Wisconsin, WEC's utilities have undergrounded 2,000 miles of overhead lines in 8 years, raising their system from 27% to 39% underground¹³. This delivered "game-changing" reliability improvements – outage minutes dropped by >97% in areas converted to underground¹⁴ – and high customer satisfaction (customers proved willing to pay for the reliability gains).
• Dominion Energy – Dominion (Virginia) initiated a $2 billion Strategic Underground Program in 2014, targeting 4,000 miles of the most outage-prone tap lines by 2028¹⁵. The goal is faster storm restoration by hardening laterals. Early results showed storm restoration times 50% faster in converted areas¹⁶. A study projected Dominion's undergrounding yields $1.76 in avoided economic losses per $1 spent, thanks to reduced outage duration (Virginia regulators valued GDP loss at ~$35k per hour per 1,000 customers)¹⁷.
• Other Leaders – Numerous public power and municipal utilities already have predominantly underground systems. For instance, Fort Collins, CO is ~99% underground with 99.9% reliability, and Colorado Springs is ~77% underground (achieved since the 1970s) with 99.9% reliability¹⁸. Anaheim, CA has been systematically undergrounding since 1990 with great success¹⁹. These early adopters underscore that 50% underground by 2040 is an achievable goal according to industry advocates²⁰²¹. Major IOUs that are now expanding undergrounding (in addition to those above) include Southern California Edison, San Diego Gas & Electric, and others focusing on high fire-risk or urban areas.

1.1 Canadian Perspective

Canada's grid remains more overhead-centric overall, with roughly only 10–12% of distribution lines underground⁴. In Québec, for example, ~11% of distribution is underground, a rate "consistent with elsewhere in Canada"⁴. This lower penetration reflects Canada's vast rural networks and historically higher cost of underground construction. That said, urban centers and new subdivisions increasingly use underground distribution for aesthetics and reliability²². Provincial programs are generally less formalized than in the US (many decisions are left to municipalities)²³, but the trend is slowly upward. For instance, Hydro Québec notes "more and more" towns choose to bury lines in new developments for resilience and beautification²². Toronto and other large cities have significant downtown underground networks (by necessity), and Canadian utilities are certainly aware of the reliability benefits. As extreme weather (ice storms, high winds) continue to impact overhead lines, interest in strategic undergrounding is growing. However, costs remain a barrier; installing underground distribution in Canada can be up to 10× the cost of overhead in some cases²⁴.

2 · Overhead vs Underground Infrastructure — Current Stats and Trends

Table 2.1 · Approximate Overhead vs. Underground Distribution Line Lengths. Sources: DOE Grid Deployment Office², Hydro-Québec⁴. (Note: Transmission lines are overwhelmingly overhead in both countries, with <1% underground at high voltages⁵.)

2.1 Distribution Systems

About one-fifth of U.S. distribution line mileage is underground, versus roughly one-tenth in Canada. In absolute terms, the U.S. has on the order of 1 million miles of underground distribution cables (out of ~5.5M total distribution line-miles)¹². Canada's total distribution network is smaller (~572,000 km or 355,000 miles³) with perhaps ~60,000 km underground. Notably, states with long-running undergrounding policies have higher shares – e.g. California ~33% underground statewide²⁵, Florida ~40% (and climbing), etc. Meanwhile, many new residential developments across North America have been built with underground utilities since the 1970s, so the underground share grows steadily as old overhead lines are retired or converted²⁶. DOE data confirms a modest upward trend (18% underground in 2009 to ~20% in 2023 nationally)². This is expected to accelerate as the large programs (PG&E, FPL, Dominion, etc.) ramp up in the coming decade.

2.2 Transmission Systems

High-voltage transmission lines remain overwhelmingly overhead in both the US and Canada. Due to much higher costs and technical complexity, only special situations warrant underground transmission (e.g. dense urban cores, submarine cable links, or environmentally sensitive areas). As of 2009, as little as 0.5% of U.S. transmission line mileage ≥200 kV was underground⁵. Industry estimates for all transmission voltages put underground at well under 5% of mileage. Canada is similar – the vast majority of its ~160,000 km of transmission is overhead²⁷, with just a few underground cable segments (for example, short runs in downtown Montreal or Toronto). The economics and repair difficulty of underground transmission mean it will remain rare, except for HVDC specialty projects or short urban sections. Thus, the focus for underground analytics is squarely on distribution networks, where underground circuits are far more common and growing.

2.3 Key Trends

In summary, North America's electric grid is still largely overhead, but the portion that is underground (millions of cable-miles) is significant and rising. Drivers for undergrounding include: resilience to storms (hurricanes, wildfires, ice), reduced routine maintenance (no tree trimming), improved public safety (no live wires falling), and aesthetics/property value gains²⁸²⁹. On the other hand, underground lines have higher upfront cost and different failure modes that can make outages harder to locate and repair. The next section discusses how utilities currently manage these underground assets and the unique challenges involved, which set the stage for why advanced analytics are needed.

3 · Managing Underground Infrastructure — Organization, Budgets, Challenges

3.1 Utility Organizational Approach

Management of underground distribution infrastructure typically involves specialized teams within a utility's operations and engineering groups. Most utilities have a Distribution Operations department that oversees both overhead and underground networks, but they often assign underground-specific crews and engineers for certain tasks. For example, underground line crews / cable splicers handle cable installation, splicing, and repairs, distinct from overhead linemen. In asset management terms, some utilities have dedicated Underground Asset Managers or Engineers focusing on cable systems, while others fold this into broader distribution asset management roles. Functional groups commonly involved include:

• Asset Management / Planning: sets maintenance strategies and replacement budgets for underground cables, transformers, and related gear. They develop asset health indices and replacement schedules for aging cable systems.
• Operations & Maintenance (O&M): field crews that perform routine inspections (e.g. manhole inspections in urban networks) and respond to underground faults. These crews require specialized training for confined spaces and cable splicing.
• Protection / Control Engineering: in urban networks, they manage network protectors and automation in underground secondary networks. In suburban networks, they oversee pad-mounted switchgear, reclosers or sectionalizers on underground laterals.
• Construction / Project teams: who execute capital projects to replace or expand underground circuits (often coordinating civil works like trenching, boring, duct bank construction).

3.2 Budget Profile

Budget-wise, underground infrastructure tends to have a different expense profile than overhead. Overhead lines incur substantial ongoing O&M costs (pole inspections, vegetation management, etc.), whereas underground lines avoid most of those but can be costly when failures occur. On average, maintenance costs for underground lines are 75–80% lower than for overhead (since there's no tree trimming or storm repair most years)²⁸. Utilities save on vegetation management – studies show undergrounding saves about $7,000 per mile per year in avoided tree trimming costs on average³⁰. Consequently, some utilities allocate a smaller routine O&M budget per mile for underground. However, they must budget for event-driven repairs and capital replacement. When an underground cable fails, repair costs can be 3–4 times higher than an overhead fix, due to the excavation, splicing, and often longer outage restoration time³¹³². Many utilities maintain contingency funds or higher capital reserves for cable replacement programs (e.g. replacing vintage PILC or early XLPE cables known to be failure-prone).

3.3 Challenges vs. Overhead

Managing underground assets is notoriously more challenging in several ways:

• Fault Detection and Location: When an overhead line trips, it's often easy to visually find the cause (e.g. a downed wire or blown fuse). Underground faults are invisible – crews have a difficult time locating faults in buried cables³³. Utilities must use specialized fault-locating equipment (thumpers, TDR reflectometers) to pinpoint failure spots, which takes time. As a result, outage restoration for underground faults is often longer than for overhead³⁴. For example, a feeder fault in a buried cable might require hours or days to locate and repair, whereas an overhead equivalent could be found and fixed in a few hours. This longer downtime is a trade-off: underground circuits fail less often, but when they do, customers may be out longer.
• Repair Complexity and Safety: Underground repairs require excavation or confined-space entry into manholes/vaults. This is labor-intensive and can be hazardous. There are safety issues with underground vaults, including risk of explosions or fires from cable faults³⁵. Sadly, incidents of manhole explosions (from arcing faults igniting gases) are known in big cities – Con Edison in NYC experiences ~2,000 smoking or exploding manhole events annually³⁶. Protecting crews and the public during these events is a major concern. Any digital solution must help prevent such incidents or at least detect them early.
• Asset Visibility: A saying in the industry is "Out of sight, out of mind." Utilities often have less granular knowledge of their underground assets' condition because they cannot be easily observed. Overhead lines get visual inspections by drone or line patrol; underground cables are hidden until failure. There is no routine visual inspection of most underground cable runs³⁷. Utilities rely on indirect indicators (load history, partial discharge tests, etc.) or scheduled tests on sample segments. This limited visibility can lead to run-to-failure maintenance more often than desired. A digital twin or monitoring solution is needed to bring underground assets "to light" in the control room.
• Data and Documentation: Managing the mapping and records of underground installations is complex. Utilities must keep track of duct banks, conduit routes, cable sizes, splice locations, etc. Historically this was paper-based. If records are incomplete, locating cables for repair or replacement becomes difficult. Modern GIS-based tools (like Schneider's ArcFM Conduit Manager) help maintain detailed maps of underground facilities, allowing utilities to know exactly which cable is in which duct, and where manholes and splices are located³⁸³⁹. Still, data quality is an issue – many utilities have discovered discrepancies in their underground network maps, which can impede timely repairs or upgrades.

3.4 Budget Implications

Due to the above challenges, utilities often end up spending a significant portion of distribution O&M and capital budgets on underground cable programs. For instance, a utility might dedicate (hypothetically) 30% of its distribution capital to cable replacements even though those cables are only 20% of the network mileage – reflecting their high criticality and replacement cost. Indeed, U.S. utility capex data shows spending on underground distribution has grown slightly faster than overall distribution spending: from ~21.4% of distribution capex in 2010 up to ~22.2% by 2019⁴⁰. This indicates utilities are investing more in underground assets (new and existing) over time.

In summary, managing underground distribution is a game of fewer, but costlier, failures. Utilities save on routine maintenance but face expensive interventions when things go wrong. The difficulty in monitoring and the high stakes of failure (in terms of outage impact and repair cost) make advanced analytics and monitoring essential. The next sections will quantify the assets at play and examine how data is (and isn't) being collected today from underground systems.

4 · Key Underground Distribution Assets in the US and Canada

The underground distribution system comprises several asset classes, each of which exists in huge numbers across North America. Below we identify the key assets and provide quantitative context:

4.1 Medium-Voltage Underground Cables

These are the primary distribution feeders and laterals in underground areas, typically carrying 5 kV to 35 kV. In the U.S., there are on the order of 1 million circuit-miles of MV underground cables (as noted, ~20% of ~5.5M total distribution miles)². Canada likely has on the order of 60–70,000 km of MV underground cable (given ~11% of ~572K km)⁴. These cables may be directly buried or in conduit. Common types include cross-linked polyethylene (XLPE) insulated cables in newer installations, and paper-insulated lead-covered (PILC) or older polyethylene in legacy systems. Many urban utilities also have extensive secondary network cables (low-voltage) in downtown grids – e.g. Con Edison (NY) alone has ~90,000 miles of underground cable if one counts all secondary, service, and primary cables⁴¹. This number illustrates that the aggregate length of underground conductors is enormous, when secondary/service cables are included (each building service adds length). Managing this vast network is a data challenge without digital tools.

4.2 Underground Distribution Transformers

These include pad-mounted transformers in suburban systems (usually in a green metal cabinet on a concrete pad) and subsurface transformers in vaults for urban areas. The U.S. has an estimated 60–80 million distribution transformers in service⁴²⁴³, of which a significant fraction are pad-mounted or below-grade units. While exact breakdown is not readily available, one industry source notes "most of the estimated 45+ million U.S. distribution transformers are overhead single-phase", implying perhaps 10–20 million are pad-mounted (underground type)⁴⁴⁴³. Each underground residential development has one pad-mount per a cluster of homes; urban network transformers each serve a city block or building. Canada likely has on the order of a few million distribution transformers (roughly proportional to population). For example, Manitoba Hydro (a mid-size utility) alone has ~75,000 km of distribution lines and tens of thousands of transformers. The key point is that millions of transformers are in underground or pad-mount configuration, which means monitoring their loading and condition (oil temperature, etc.) is often more challenging (no pole for a visual inspection). New IoT sensors are emerging to monitor these sealed transformers remotely.

4.3 Underground Switchgear and Protective Devices

These include underground switches, sectionalizers, and network protectors. Pad-mounted switchgear (in cabinets) is common for 3-phase underground feeder loops in subdivisions – allowing utilities to sectionalize faults. In urban networks, network protectors are installed on secondary network transformers; a city like New York has hundreds of network protectors that require monitoring. These devices often have SCADA or remote telemetry in modern systems (for example, network protectors report status over fiber). Fault indicators are another asset – on long underground feeders, utilities install faulted circuit indicators (FCIs) in manholes or pad-mounted enclosures to help locate faults by showing a tripped flag or sending a signal. Digital sensors are replacing simple FCIs, as discussed later. The quantity: large utilities may have thousands of underground switching points and FCIs deployed.

4.4 Manholes and Vaults

These are the access points to underground cables and equipment, especially in urban areas with "spiderweb" secondary networks. The number of manholes/vaults is enormous in major cities. For example, Con Edison's underground system in NYC has ~267,000 manholes and service boxes⁴⁵. Each contains cable splices, joints, or transformers. Other cities (Chicago, Boston, etc.) have tens of thousands of utility manholes each. Even smaller cities with downtown underground districts will have hundreds to thousands. These structures present maintenance challenges (water pumping, stray voltage hazards, etc.) but also opportunities for sensor installation (one can place environmental sensors in a manhole, as Con Ed has done). From an analytics perspective, tracking asset data by location (manhole ID) is important. Modern GIS-based asset systems (like Schneider's ArcFM Conduit/Structure Manager) explicitly model these structures and what's inside them⁴⁶. This is crucial for map-based visualization of cable health – you need to know which segment is between which vaults.

4.5 Cable Accessories

Finally, numerous smaller components like splices / joints, terminations, and elbows (in pad-mount gear) are key parts of underground systems. Failures often occur at these points (a bad splice or terminator can overheat or allow moisture). A typical underground feeder might have dozens of splices. Utilities keep inventories of splice types and ages, since certain older splice designs are known to be problematic. Any AI solution should consider accessory data, though utilities historically have fewer records on them. Modern digital partial discharge sensors can sometimes detect discharge at specific splices, identifying a failing joint. Quantitatively, if there are ~1 million miles of underground cable in the US, there are likely tens of millions of splices (assuming an average cable segment length of a few hundred feet between splices).

5 · Data Collection and Monitoring of Underground Infrastructure Today

Figure 5.1 · Today vs Tomorrow The left side shows what a utility sees today — a flat GIS map of cable routes. The right side overlays the Earthflow Physics AI™ risk score on the same network. The data layer is the binding constraint, not the cable itself.  ·  See it live →

Utilities traditionally rely on a combination of SCADA systems, periodic testing, and reactive monitoring to manage underground equipment. Below we outline current practices and their limitations:

5.1 SCADA and Remote Monitoring

For the most part, distribution-level SCADA coverage on underground feeders is limited. Unlike transmission lines which have extensive metering, most distribution cables do not have real-time sensors along their length. However, SCADA is present at substations feeding underground circuits (measuring feeder currents, breaker status, etc.), and at some key points in the network: e.g., pad-mounted reclosers or switches may be SCADA-controlled, and network protectors in urban grids often report status remotely. SCADA gives macro-level data – if a feeder breaker trips, the control center knows it and can see the current/voltage to the fault point. But SCADA usually cannot pinpoint where along an underground feeder the fault is, nor give granular condition data. Some utilities have added smart fault indicators that communicate back to SCADA or an AMI network. For instance, there are underground fault sensors (by SEL, Sentient, etc.) that can wirelessly report fault passage and even load data from inside a pad-mounted transformer or vault. These are typically integrated with Distribution Automation systems (DA or ADMS). As an example, Sentient Energy's UM3+ smart underground sensor can monitor up to 12 cable phases in a vault or switch cabinet, measuring load and fault waveforms and sending data via cellular or mesh network⁴⁷⁴⁸. Such sensors effectively extend SCADA visibility into the underground domain, capturing high-resolution current and voltage data at points along the feeder. Currently, deployment of these is growing but not ubiquitous – usually part of targeted reliability projects.

5.2 Outage Detection and Customer Calls

In many distribution systems, the first "sensor" that an underground cable failed is the customers. Lights go out, customers call in or (with smart meters) the AMI system flags a loss of power. An Outage Management System (OMS) then deduces the likely failed component. Smart meters have improved this process by pinging meters – if an entire segment of meters is dark, the utility infers which cable section is out. This is a reactive approach and doesn't give condition info prior to failure, but it is how many distribution outages are initially identified. Some utilities have also deployed "last gasp" devices: certain pad-mounted transformers or network protectors send an alert when they lose power, aiding in outage localization. Again, this is event-driven data, not continuous monitoring.

5.3 Periodic Field Inspections

Because you cannot see an underground cable directly, utilities use testing technologies. Common practice includes infrared (IR) thermography on accessible terminations (e.g. open a transformer cabinet and IR scan the cable elbow for hot spots), cable insulation testing (offline tests such as VLF tan-delta or offline partial discharge on de-energized cables), and online partial discharge surveys (attaching sensors to detect PD while cable is live). Some utilities have proactive programs: for example, testing a sample of older cables each year to assess health. There are also cable monitoring vans that can inject signals to measure impedance changes, etc. However, these tests are labor-intensive and typically done infrequently (maybe a cycle of every few years for a set of cables). IMCORP, a specialty firm, has technology to test new and service-aged cables to find defects and verify integrity, and they report that only ~60% of new cable installations meet factory standards upon commissioning (meaning 40% have some defect that could shorten life)⁴⁹⁵⁰. Regular testing can find such issues and "reset the clock" by fixing them⁵¹. Still, many utilities struggle to allocate enough budget and outages to test a large percentage of their cables, so these programs often cover only the highest-risk segments.

5.4 Environmental and Condition Sensors

A newer development is the installation of IoT sensors in underground structures to continuously monitor conditions. A prime example is Con Edison's deployment of manhole sensors in New York. Con Ed has worked with a vendor (CNIGuard) to put multi-sensor IoT devices in manholes, measuring parameters like combustible gas levels, stray voltage, humidity, flooding, and temperature⁵². These sensors report via wireless links to a cloud analytics platform, which can alert the utility if, say, a manhole is heating up or filled with gas – conditions that precede manhole fires or explosions⁵³. The system even includes IR cameras that can detect cable "hot spots" so crews know exactly where to intervene⁵⁴. Con Ed began deploying these in 2017 and has been expanding these to cover more of their 266,000 manholes⁵⁵. This is a great example of modern IoT data collection for underground assets. Similarly, some utilities install temperature and oil-level sensors on underground transformers (especially large network transformers) to catch overload or oil leaks. Distributed Temperature Sensing (DTS) is another technology: a fiber run alongside a high-voltage cable can detect temperature changes along its length, effectively turning the cable into a sensor for hotspots – this is more used in transmission cable tunnels but could be applied to distribution feeders in long ducts.

5.5 Fault Indicators and Line Sensors

Many underground systems make use of faulted circuit indicators (FCIs) that are either manually checked or communicate remotely. Traditional FCIs are simple devices that clamp on a cable and pop a flag or light when a fault current passes. Newer intelligent FCIs (line sensors) do much more – for example, SEL's underground sensors or Sentient's UM3+ can capture high-resolution current waveforms, detect disturbances, and time-stamp and geo-locate fault events⁴⁸⁵⁶. They can differentiate transient vs permanent faults and help estimate distance to fault by comparing waveforms at multiple sensors. These are typically installed at cable sectionalizing points (switches, transformer feed-thrus). They significantly speed up fault isolation – rather than send a crew to manually dig up multiple splice pits, the control center can see which sensor saw the fault and which did not, narrowing down the fault section immediately. A cited result: with such sensors and analytics, utilities have achieved 20%+ improvements in Customer Minutes Interrupted (CMI) and up to 65% reduction in patrol time to locate faults⁵⁷⁵⁸. This not only improves reliability metrics but also cuts labor costs (patrol crews driving around for hours). We will detail these ROI figures later, but it underscores how adding even basic sensors yields big benefits.

5.6 Gaps and Needs

Despite the above tools, there are still gaps in data collection for underground assets:

• Real-time condition monitoring is rare outside of pilot projects. Continuous online partial discharge (PD) monitoring, for instance, is not widely deployed in distribution (though it's used in some substation cables). Companies like Omicron offer permanent PD monitors for cables⁵⁹, which can alert if insulation degradation reaches critical levels, but few utilities have blanketed their network with such devices due to cost.
• Integration of disparate data: The data that does exist (from SCADA, outage records, test results, sensor pilots) is often siloed. A comprehensive analytics solution needs to pull all these together – e.g., combine the SCADA/OMS outage history with any field test data and the GIS asset data to see patterns.
• Sensor coverage: Many utilities simply lack sensors on the majority of their underground feeders. For example, a typical medium-sized utility might only have SCADA at the substation feeder breaker and maybe halfway down a feeder at a normally-open tie switch. Everything in between is dark until a failure. Expanding sensor coverage (with devices like Sentient's UM3+ or others) is a need, but it must be cost-justified.

Before discussing KPIs (Chapter 7) and emerging sensor hardware (Chapter 8), we first need to confront an uncomfortable reality: the data that does exist on the underground side is often of materially worse quality than what utilities are accustomed to working with on the transmission side — and the volume of data emerging from modern AMI and sensor deployments compounds the problem. The next chapter examines both.

6 · Data Quality & Volume Realities

An honest white paper on underground distribution analytics cannot skip this chapter. The biggest gap between an idealized AI/ML solution and a deployable one is not the algorithm — it is the state of the data the algorithm has to consume. This chapter names the specific data-quality problems on the distribution and especially the underground distribution side, the volume problems that emerge when you fuse modern AMI and sensor streams, and the design choices a credible solution must make.

6.1 The Honest Premise — Distribution Data Is Materially Worse Than Transmission Data

Transmission grew up with SCADA-grade telemetry on every line. Synchrophasors, digital relays, line monitors, and decades of disciplined record-keeping have produced data assets that are largely complete, consistent, and recently calibrated. The distribution grid evolved bottom-up from CIS billing systems and paper switching diagrams, and its data layer reflects that history: heterogeneous, partial, and frequently stale.

Underground specifically suffers more than overhead distribution for one structural reason — you cannot inspect what you cannot see. Overhead lines get visual pole patrols and increasingly drone surveys; those inspections continuously correct the asset record. Underground cable runs are last visited when they were installed (in some cases, 60+ years ago) and the record from that installation may or may not have survived. The ground-truth feedback loop that keeps overhead GIS clean is largely absent below grade.

6.2 Data-Quality Issues by Source System

The problems are specific to each source. Below is what utility analytics teams typically encounter when they try to build a fused dataset for underground cable analytics:

Table 6.1 · Data-quality issues typical of each source system feeding an underground cable analytics platform.

6.3 The Volume Problem — What Happens When You Fuse Everything

Set the quality problem aside for a moment. Even if every source were clean, the volume of data that a fused analytics platform has to manage is substantial — and growing fast as utilities deploy more sensors:

• AMI: 15-min interval data × millions of meters × multi-year retention = multiple TB per utility per year. Most utilities retain only daily or hourly summary aggregates; raw interval data is huge.
• Line sensors (Sentient UM3+ class): Continuous 256-samples-per-cycle waveform capture per device. A pilot of 50 sensors generates on the order of 1 GB of raw waveform data per day. Multiply by a feeder-wide deployment.
• Distributed Temperature Sensing (DTS): One temperature reading every meter along the fiber every few seconds. A single 5-mile fiber-traced feeder produces GB per day.
• Online partial-discharge monitors: Pulse-count time series with severity classification — not large per device, but high cardinality across a fleet of cable terminations.
• SCADA / historian: Decades of historical tags in PI Historian or equivalent. Sub-minute SCADA across a utility's substation fleet accumulates to billions of rows over a 10-year window.
• Image and IR data: Manhole IR camera images, drone inspection photos, satellite imagery for routing and terrain. Unstructured, high-bandwidth.
• Cross-system joins: Joining GIS + OMS + SCADA + AMI + sensor data at the per-segment level is a wide many-to-many. A single utility's fully-joined dataset can run to billions of rows.

The volume problem is not just storage. It is making the join coherent at the cable-segment level, with confidence metadata about which source contributed what to each output. A modern utility analytics stack has to be designed for this from day one, not retro-fitted.

6.4 The Underground-Specific Compounding Factors

Four factors compound the data problem specifically for underground — they are less severe on the overhead side:

• No visual inspection. The ground-truth feedback loop that corrects bad GIS records on the overhead side is absent below grade. Errors persist for decades.
• Slower observation cycle. Underground failures happen 3–5× less often than overhead. The supervised-learning signal that an ML model trains on is correspondingly sparser, which is both good (fewer failures, what we want) and bad (less labeled training data).
• Higher cost per observation. A test on an underground cable typically requires de-energization, sometimes excavation, and mobilizing a specialized splicing crew. Utilities can afford to test only the highest-risk segments — biasing the test-result dataset toward the worst cables.
• Records gap from undergrounding waves. Cables from the 1970s suburban-development boom often have the worst documentation because record-keeping standards at the time were primitive. These are also among the oldest cables in service today — the segment with the most degradation has the worst data.

6.5 What a Realistic Solution Must Do

The implication for any AI/ML solution for underground cable analytics: you cannot design for a clean-data utopia and call it a day. A solution that works in production has to make five design choices up front:

1. Tolerate missing data. Industry-default imputation when utility-specific data is missing or stale, with explicit confidence intervals on outputs. A health-index score with a wide confidence band is more useful than a precise number that pretends data it doesn't have.
2. Provenance per output. Every numeric result traces to which source contributed what — so regulators, engineers, and rate-case reviewers can audit the basis for any specific recommendation. This is non-negotiable when the output of the system drives capex decisions defensible to FERC and state PUCs.
3. Disclose data-quality caveats in answers. When the analytics layer (and especially an agentic-AI Q&A layer) answers a question, it should surface the data-quality caveats. "AHI score 4.2 for segment X is based on GIS install date and SSURGO soil data — no test results on file in the last 8 years" is a more honest answer than just "4.2".
4. Volume-aware architecture. Cloud-scale processing for the AMI, sensor, and historian volumes described above. Batch + stream patterns where each is appropriate. Segment-level join indices that don't require shuffling billions of rows for every query.
5. Schema-flexible ingest. Auto-mapping across the common utility GIS, OMS, and SCADA schemas (Esri ArcFM, Schneider Conduit Manager, generic geodatabases) with manual override for the long tail of in-house schemas. The Schema Auto-Mapper is the single most important architecture decision for 7-day deployment.

Next, given these realities, we describe the KPIs and health metrics utilities should still track — with explicit acknowledgment of which KPIs are robust to messy data and which require investment to make them reliable.

7 · Key KPIs for Underground Cable Health and Risk

Evaluating the health and remaining life of underground cables requires a combination of reliability metrics and asset condition indicators. Utilities and industry research have developed several Key Performance Indicators (KPIs) and indices for this purpose:

6.1 Failure Rate and History

The most basic metric is the failure rate of cables, often expressed as failures per 100 miles per year. Utilities track how often their underground cable segments fail (e.g., 1 fault per 20 miles per year would be a rate of 5 per 100 miles/year). They also look at historical failure count per cable or per circuit. A cable that has failed multiple times (with splices each time) is at higher risk for future failure. Past failure history is a strong indicator in health indexing – if a particular vintage of cable shows, say, 0.2 failures/mi-year vs another type at 0.05, that informs replacement prioritization. Some utilities incorporate a "cable fault history score" into their asset health index (e.g., score = 5 if cable had >2 failures in last 10 years, etc.).

6.2 Cable Age vs. Expected Life

Age (time in service) is a simple but important KPI. Many utilities assume a nominal design life of ~40 years for underground distribution cables⁶⁰, though actual life can be much longer if conditions are good (there are cables 60–70 years old still in service). Age by itself doesn't equal poor health, but age combined with certain insulation types is telling. For example, early-generation XLPE cables from the 1970s often had issues and many failed in the 20–30 year range due to insulation deterioration. Likewise, paper-lead cables might last a long time but become riskier if the lead sheath is degraded. Utilities often create an Age Score (older = higher risk) as part of a health index. If combined with failure stats, one might find, for instance, that cables beyond 30 years old have failure rates double those under 30 years, etc. Age is an easy KPI since utilities have installation dates in their databases for most cables.

6.3 Insulation Type and Condition Tests

The type of cable and its condition indicators form another KPI set. For instance:

• Insulation Material (PILC, XLPE, EPR, etc.) – Different materials have different typical failure modes. PILC (paper insulated, lead covered) cables can last a very long time but are susceptible to moisture ingress if the lead sheath is damaged. XLPE can suffer "treeing" (water tree degradation) especially in older formulations. EPR (rubber) cables handle moisture better but have other aging characteristics. Utilities may assign higher risk to known problematic types (e.g., "pre-1985 XLPE cable" might be flagged due to known issues with that era's manufacturing).
• Test Results (Tan Delta, Partial Discharge) – If a cable has undergone diagnostic testing, those results become KPIs. Tan delta (dissipation factor) testing reveals the insulation dielectric losses – a high tan delta indicates deteriorated insulation. Partial discharge (PD) test or monitoring results are critical: measurable PD activity in a cable or accessory is a red flag for insulation defects. A cable health index might incorporate, for example, PD severity level (none, low, high) as a KPI. Online PD monitors can even quantify PD in picoCoulombs or give a trend of PD event counts. Some utilities perform offline PD after cable installation – as IMCORP data showed, ~40% of new cables had detectable PD or defects that needed correction⁶¹⁵⁰. Catching those early greatly reduces future failures. For in-service cables, a spike in PD activity is a predictor of failure within months or a few years if not addressed.

6.4 Load / Stress Indicators

The operating conditions of the cable serve as KPIs for remaining life:

• Load Factor / Peak Load: Cables that run near their capacity (high ampacity usage) run hotter, which accelerates insulation aging. A cable consistently loaded at (say) >80% of rated current during summers might have a shorter life. Thus, tracking the peak load each year or the average load factor is useful. Some utilities compute a "stress index" combining load factor with cable size – essentially how hard the cable is working relative to its design.
• Thermal Events: Any history of overloads or high-temperature alarms (if any sensor or IR reading) would be a KPI. If an underground cable ever exceeded its temp rating (perhaps during a contingency when it picked up another feeder's load), that thermal excursion can cause cumulative damage.
• Environment: External environment KPIs include soil conditions (corrosive soil can degrade lead sheath), presence of water (submerged ducts lead to more rapid deterioration unless cable is fully waterproof design), and ambient temperature. Utilities might not have real-time data on this, but they know, for example, if a cable is in a dry sandy soil versus a wet clay. Some even factor a flooding risk index – cables in areas prone to flooding (or in ducts known to fill with water) may be given a higher risk score.

6.5 Reliability Impact Metrics

At a higher level, utilities also consider how a cable's failure would impact the system – effectively a criticality KPI. For example:

• Customers Served / Load Criticality: A feeder cable serving 1,000 customers or a major industrial load has higher consequence of failure than a lateral serving 50 homes. So "customers affected" is a KPI for risk (failure risk × impact).
• SAIDI / SAIFI Contribution: Some utilities retrospectively measure how much each feeder or cable contributed to reliability indices. If a particular underground feeder has contributed a lot to SAIDI due to long repairs, it might get priority for upgrade. Conversely, areas where overhead outages dominate SAIDI might shift focus away from underground. However, since underground outages are rarer but longer, they can disproportionately affect CAIDI (restoration time). The System Average Interruption Duration Index (SAIDI) typically improves as undergrounding increases – one study found a 10% increase in underground line share corresponded to ~14% reduction in SAIDI (including major events)⁶². Thus, tracking SAIDI for underground vs overhead areas is another performance metric.

6.6 Asset Health Index (AHI)

Many utilities synthesize the above factors into an Underground Cable Health Index, often a score from 1 (new/good) to 5 (end-of-life poor), or similar. This AHI is a composite KPI, weighing age, type, test results, failure history, etc. For instance, a possible scheme:

• Score 1 (Excellent): New XLPE cable <10 years, no failures, low load.
• Score 3 (Fair): 30-year-old cable, one past splice repair, moderate PD readings.
• Score 5 (Poor): 40+ year cable, multiple faults, high leakage or PD, frequently loaded near capacity. This is analogous to health indices used for transformers or other assets. The AHI allows fleet-wide comparison to prioritize replacements. A cable with health index 5 would be a strong candidate for proactive replacement before it fails again.

Figure 6.1 · Decision Matrix Per-segment AHI 1-5, RUL bands, and recommended action: REPLACE / REPAIR / RETIRE / DEFER / MAINTAIN. Each cell carries provenance JSON pointing back to the source datasets used.

6.7 Remaining Useful Life (RUL) Estimation

Some advanced analytics attempt to directly predict remaining life in years. This often piggybacks on the health index – essentially translating the HI and degradation trends into a probabilistic RUL. For example, if PD activity is trending up, an algorithm might predict failure likely in <2 years, giving an estimated RUL distribution. While exact RUL prediction is tricky, it's an ultimate KPI the industry is interested in: knowing whether a given cable likely has 5, 10, or 20 years left can inform capital planning.

In practice, utilities combine leading indicators (like PD, test results) with lagging indicators (failure history) to manage cable risk. The KPIs above feed into decisions like: which cables get replaced first, which ones need monitoring, and how to budget for renewals. The challenge is having the data for these KPIs – which leads us to what an ideal digital solution would ingest and analyze.

8 · Examples of Sensor Hardware Solutions and Their Data

Several specialized sensor and monitoring solutions exist to gather data from underground electrical infrastructure. Below are some notable examples and the data they provide:

7.1 Smart Underground Line Sensors (Fault and Load Monitors)

One example is the Sentient Energy UM3+ underground line monitor. This device can be installed in pad-mounted switchgear, junction boxes, or even submersible vaults (it has a compact, submersible form factor for manholes)⁷⁴. The UM3+ clamps around cable conductors (up to 12 phases with one unit) and measures:

• Fault detection and indication: It instantly detects fault currents on the cable and can timestamp and communicate an alarm⁶⁴. This helps pinpoint faulted sections by seeing which sensors triggered.
• Load and current measurement: It provides highly accurate load current readings and can capture the magnitude of fault current as well⁷⁵. This allows both real-time load monitoring and post-fault analysis of how severe a fault was.
• Waveform capture: Importantly, the UM3+ can record high-resolution current waveforms (256 samples per cycle) for disturbances⁷⁶. This means it can capture the oscillography of a fault or transient. Those waveforms are sent to the cloud (or utility analytics platform) for analysis of fault characteristics or precursors.
• Time synchronization & location: It has GPS for location and time syncing⁷⁷. This allows multiple sensors' data to be correlated accurately (essential for triangulating fault location by time-of-arrival differences, or for aligning waveforms from different points).
• Communications: It supports cellular or mesh communications and can be configured remotely⁷⁸. Data is fed into an "Ample" analytics software or SCADA.

Data Provided: The utility receives real-time alerts (trip signals), numerical load data (currents/voltages if equipped for voltage), waveform files for events, and periodic status. With analytics, one utility reported these sensors led to 20%+ reduction in feeder outage minutes and 65% reduction in patrol time, as noted earlier⁷⁹. They also achieved an 11% O&M cost reduction due to efficiency gains from sensors⁵⁸. These numbers show how valuable the data from such sensors can be.

7.2 Manhole / Vault Monitors

As described with Con Edison's case, a product by CNI Guard (called Underground Sentinel or similar) is used. These multi-sensor IoT devices installed in manholes typically measure:

• Combustible gas concentration: detecting gas buildup (like methane, carbon monoxide) from smoldering insulation or sewer gas, to preempt explosions⁵².
• Atmospheric conditions: temperature and humidity in the manhole, which can indicate overheating cables or flood conditions⁵².
• Salinity or water presence: some sensors detect presence of salt water (road salt ingress can cause conductive paths leading to faults)⁵².
• Stray voltage: detecting any voltage present on manhole covers or surrounding surfaces (important for public safety – energized structures can occur if insulation fails).
• Cameras (IR and optical): Con Ed's units include an infrared camera to spot hot spots on cables/connectors and a regular camera for images⁸⁰.
• Accelerometer or tilt (optional): to detect if a manhole cover dislodges in an explosion.

Data Provided: These monitors typically send continuous telemetry. Con Ed's system transmits data to a dashboard and sends alerts if any parameter goes out of bounds⁸¹⁸². For example, if a cable in a manhole starts arcing, the sensor could detect rapid temperature rise and a spike in carbon monoxide, triggering an immediate alert to dispatch crews before a major fire or outage occurs⁵²⁸³. The data can be integrated into the analytics platform to show a map of manhole conditions. Con Ed's initial deployment has prevented incidents by indicating cables "in need of attention or repair" before they fail⁵⁵. A value proposition from their experience: the IoT sensors act as a "virtual inspection team," reducing the need for physical inspections of hundreds of thousands of structures⁸³.

7.3 Partial Discharge (PD) Monitoring Systems

For higher-voltage cables (and applicable to distribution too), there are systems like OMICRON MonCablo or similar by other vendors. These install coupling sensors (capacitive or inductive) usually at cable ends (terminations or switchgear) to continuously listen for PD pulses. They can cover many kilometers of cable by using time-of-flight of pulses to localize PD sources along the cable.

Data Provided: They output PD magnitude, pulse count, phase-resolved patterns (to distinguish between different defect types), and trending over time⁵⁹. These systems often have software that can alarm if PD exceeds thresholds. Utilities might deploy these on critical feeders or after a suspicious test result. If our AI platform receives PD data, it would treat increasing PD as a key indicator to take action (for instance, schedule a section replacement at a detected PD site).

7.4 Distributed Temperature Sensing (DTS)

If an optical fiber is embedded in or alongside a cable, DTS equipment can send pulses of light and measure backscatter to determine temperature every meter along the fiber.

Data Provided: Essentially a temperature profile of the cable route in real time. This can identify localized heating (perhaps due to a bad splice joint or an overloaded segment in a duct section with poor heat dissipation). While more common in transmission cables, distribution could use it where fiber is available (some newer cables or duct installations include fiber conduits). Real-time thermal data would feed into dynamic rating calculations (increasing utilization safely) and alert if something is running hotter than normal.

7.5 Smart Cable Bolts / Termination Sensors

A niche but interesting category: sensors embedded in cable connectors or terminations. An example: some companies offer wireless temperature sensors that attach to cable lugs or elbows to monitor connection temperature (E.g., Eaton or ABB have devices to monitor bolted connections in switchgear). These are useful because a loose connection in a pad-mounted transformer will get hot and eventually fail.

Data Provided: Point temperature readings of critical connection points. These often communicate via low-power radio to a local gateway. A spike in temperature at a particular elbow might indicate an incipient failure (or just an overload).

7.6 Advanced Fault Indicators (Stand-alone)

Companies like SEL and Horstmann (in Europe) provide underground fault indicators that can detect both short-circuit and earth-fault currents continuously⁸⁴. Some models mount inside transformer cabinets and use current transformers around cables. Modern ones can communicate via cellular or radio.

Data Provided: Indication of a fault pass and sometimes load values. Not as rich as Sentient's full sensor, but if networked, they feed into the outage analysis.

Next, we discuss how one could implement an underground cable analytics solution quickly using existing data (without waiting for full sensor deployment), as well as the value proposition and ROI of doing so.

9 · Data, Sensors, and Analytics for an AI/ML-Driven Underground Cable Solution

Building a robust AI/ML-driven analytics solution for underground cable health requires integrating many data sources and sensor inputs, and presenting insights in a user-friendly (ideally map-based) interface. Here we outline the key data types, sensors, and analytics capabilities needed:

8.1 Asset Data (Inventory and Attributes)

The foundation is the utility's existing asset registry and GIS data for the underground system. This includes for each cable segment: installation date (age), length, route geometry (for mapping), voltage level, conductor size and material, insulation type (XLPE / PILC / etc.), manufacturer if known, and any construction notes (duct vs direct-buried, etc.). It also includes connectivity – which cable connects which equipment (from substation breaker down to loads). Using the GIS, the solution can map each cable segment to specific coordinates. Off-the-shelf, most utilities have this data in their GIS or asset database. A solution should readily ingest this "already available data"⁶³ – as Siemens notes in its Cable Analytics offering, analyzing underground cables based on existing utility data (age, load, etc.) can reduce failures dramatically⁶³. This implies the solution must parse GIS line data, perhaps through a utility's Esri geodatabase or similar, and assign unique IDs to each cable section.

8.2 Historical Performance Data

This includes failure/outage history for the cables. An ideal solution will interface with the utility's Outage Management System (OMS) or reliability database to pull records of past underground cable faults, including date, location (which cable or approximate), cause (if determined), and outage duration/customers affected. This allows the AI to train on what factors correlate with failures. For example, it might learn that cables of a certain vintage in wet soil had X failures. Additionally, any work order history (repairs, splice installations) is valuable data. If a cable had 3 splices, that's important to know (splices can be weak points). Some utilities track the number of joints on a cable segment – that could be a data field to include.

8.3 Load and Electrical Data

Integration with SCADA or distribution management systems is needed to get electrical measurements on the cables/circuits. Minimum data would be feeder-level loading profiles (how much current the cable sees over time). If SCADA gives only substation feeder current, the solution can assume that load is distributed, or better, use power flow in the network model to estimate each segment's loading. Ideally, AMI (smart meter) data could also be used to estimate loading on each segment by summing the meters beyond it. This gives a more granular view of how heavily each cable is loaded at peak times. For circuits with switching, the system should know the current network configuration to assign loads correctly. Additionally, any voltage or power quality data from sensors could be input. For instance, if there are power quality monitors on a feeder, the presence of voltage sags, harmonics, etc., might indicate something (e.g., repeated voltage sags could mean incipient faults). At a minimum, feeder peak currents by season and any overload events should be fed in.

8.4 Sensor and Condition Data

This is where newly deployed sensors come in:

• Intelligent Line Sensors / FCIs: Data from devices like the Sentient UM3+ underground sensor should feed the platform. These provide time-stamped fault currents, high-resolution waveforms of disturbances, and real-time load readings along the feeder⁶⁴⁶⁵. With multiple such sensors, the system can use AI to do event analytics – e.g., pinpoint fault location by comparing sensor trip timings, or detect an incipient fault by seeing small waveform anomalies (a capability mentioned by vendors for detecting things like arcing cable before it fails completely⁶⁶⁶⁷). The platform's ML algorithms can be trained on waveform patterns to classify "this looks like partial discharge" or "that looks like a momentary fault likely from dig-in". Cloud computing is often used for this due to the data volume⁶⁸. The solution should integrate with whatever communications the sensors use (DNP3 SCADA, MQTT, cellular API, etc.) to stream or periodically poll the data.
• Environmental IoT Sensors: If the utility has manhole monitors (like Con Ed's), those data (gas levels, manhole temperature, etc.) should be integrated. For example, if a sensor detects rising carbon monoxide in a vault (a sign of smoldering insulation), the analytics could flag that cable as at immediate risk. Similarly, water level sensors could update which cables are underwater. The system might show on a map that "Manhole 1234 has water to 6 inches" which correlates to cables X, Y, Z being submerged – raising their near-term failure probability.
• Partial Discharge Monitors: Should the utility install online PD monitoring devices at cable ends or substations (some install coupling capacitors to detect PD on feeders leaving a substation), that data is gold for predicting failures. The analytics engine should ingest PD event counts, magnitude, etc. If PD is concentrated at a certain location (some systems can triangulate PD location along a cable within some error), that can be mapped and trigger an alert like "Cable segment between manhole A and B showing PD – likely deteriorating insulation."
• Thermal Monitoring: If distributed temperature sensing or smart thermal sensors are present, their data is input. A continuously high temperature on a cable indicates possible overload or a failing splice (a high-resistance joint gets hot).
• Mobile Inspection Data: Data from periodic IR scans or offline tests can be stored in the system. For example, after a VLF test, the measured tan delta or leakage current can be recorded as a parameter for that cable. The AI could incorporate that into the health score. Integration with maintenance management systems can allow automatic import of test results (e.g., attach PDF of test or numeric results to cable record).

8.5 External Data (for Context)

External factors like weather and events can be relevant. For instance, the system might ingest severe weather data – if an area was flooded or experienced a lightning surge, that could stress cables. Soil maps or corrosion indices can also be input. Even traffic pattern data can be relevant for road-passed cables (vibration, though likely minor compared to electrical stress). At minimum, noting environment type (e.g. coastal salty area vs dry inland, or direct buried vs conduit) helps the ML model cluster cables by similar environment.

8.6 Analytics and AI/ML Needed

With the above data, the solution would perform several analytic functions:

• Asset Health Scoring: Calculate a real-time or regularly updated health score for each cable segment using ML models. This could be a supervised model trained on historical failures (features = age, type, load, etc., label = failed or not within X years). The model can output a probability of failure or a health score from 0–100. Siemens, for example, claims its AI-driven analysis can "reduce cable failures by up to a factor of 10" by identifying at-risk cables before they fail⁶³. This implies the model can effectively highlight the worst condition cables which can then be replaced or repaired proactively, drastically cutting the observed failure rate.
• Remaining Life Prediction: Using techniques like survival analysis or regression, the analytics can estimate RUL (with confidence intervals). If certain cables show rapidly increasing partial discharge activity, the model may predict high failure likelihood in the next year, prompting immediate action. Conversely, cables in good shape might be projected to have 10+ years left, allowing deferral of replacement (Siemens notes extending asset life by up to 10 years via these analytics⁶³).
• Anomaly Detection: The ML can continuously monitor incoming sensor data for anomalies. For example, detecting precursor events – small momentary disturbances, slight insulation leakage currents, etc. Sentient Energy's analytics modules do this to "predict and preempt outages" by spotting waveform anomalies that precede faults (like momentary current spikes from incipient arcing)⁶⁶⁶⁷. Machine learning algorithms (possibly unsupervised) would flag unusual patterns on a feeder that deviate from baseline, triggering an investigation before a full outage occurs⁶⁹⁷⁰.
• Risk Ranking and Decision Support: The solution should present a prioritized list of cables by risk (combining probability of failure with consequence like customers affected). It might simulate what happens if that cable fails (outage hours, etc.) and therefore quantify the risk in e.g. customer-minutes. Hitachi Energy's Lumada APM, for instance, includes modules for asset risk and optimization that let users "simulate intervention scenarios" and balance performance, risk, and cost⁷¹⁷². For cables, one could simulate: "What if we replace the 5 worst segments? How much SAIDI improvement vs. cost?" The AI can optimize which projects yield the best reliability gain.
• Map-Based Visualization: Given the spatial nature of distribution networks, a GIS-based visualization is critical. The platform should display the distribution circuit map with color-coded segments indicating health or risk level. Users should be able to click on a cable segment and see its data (age, load, last test result, health score, etc.). Likewise, sensor locations (like smart switches, manhole sensors) should be plotted with status icons (e.g., a manhole icon turns red if high gas detected, or a sensor icon flashes if it picked up a fault). An example is Schneider's ArcFM-based solutions – while they manage data, an analytics overlay could use the ArcFM map to show, say, duct bank occupancy and cable risk³⁸³⁹. This map-centric approach helps experts and field crews alike – they can see geographically where the hotspots are. For instance, a cluster of "red" high-risk cables in one neighborhood might suggest a systemic issue there (perhaps soil or a bad batch of cable). Map visualization also aids planning – e.g., to coordinate cable replacement with other civil works. The user request explicitly asked for map-based visualization if possible, and indeed, modern ADMS and asset tools use web-GIS interfaces to display asset condition layers.
• Integration with Workflow: The analytics solution should integrate outputs into the utility's workflow: generate work orders or inspection orders for cables flagged as high risk, send alerts to control centers if a sensor threshold is exceeded, and provide reports for regulators or reliability compliance showing improvements. It may tie into maintenance management (EAM systems) so that when the AI flags "Cable XYZ has high failure risk," a maintenance planner can seamlessly schedule a cable rejuvenation or replacement.

8.7 Example of Solution Implementation

Siemens offers a Cable Analytics solution via its Advanta services, which emphasizes using advanced data analytics on existing utility data to target cable failures. They claim it can significantly reduce failures (by up to 10×) and defer capital by 10+ years⁶³. This suggests their approach uses the data points above to identify cables that can either be treated (maybe injected with rejuvenation fluids) to extend life or replaced just-in-time instead of too early. Hitachi's Lumada APM likewise would ingest these data streams into its health and reliability models, providing a "single pane of glass" for asset condition⁷³⁷¹. The combination of physics-based models (e.g. thermal aging calculations) and AI models (pattern recognition from big data) yields a comprehensive view.

In the next section, we give concrete examples of sensor hardware available today and the data they produce, which feed the kind of analytics described.

10 · Rapid Implementation Using Off-the-Shelf Utility Data

A valuable aspect of an AI-driven cable analytics solution is that it can often be deployed rapidly using the utility's existing data, then enhanced over time with new sensors. The question asks how such a solution could be implemented with off-the-shelf data like cable age, material, voltage, etc., for quick deployment. Here's a proposed approach:

1. Leverage Existing GIS and Asset Databases: As a first step, ingest all readily available asset data on underground cables. Most utilities have GIS mapping of their network and attribute tables for equipment. By importing this, the solution immediately knows the inventory of cables and their key attributes (length, install date, type). This alone enables basic risk modeling – e.g., identifying all cables older than 40 years or all cables of a certain type that has known issues. Without any new sensor, one can apply industry failure curves for different cable types, the software can estimate a baseline failure probability for each cable based on age and type. Many utilities have done paper analyses of this; the software makes it systematic and map-based.
2. Integrate Historical Outage Records: Pull in the utility's last, say, 10 years of outage records for underground equipment. This will show which segments failed and how often. Using this, the solution's ML model can find patterns (e.g., "cables installed in the 1980s in Area X have had 3x the failure rate of average"). Off-the-shelf data here is just the utility's own outage logs. This can typically be exported from OMS or reliability reporting spreadsheets. The AI can be initially trained on this supervised dataset (features = cable attributes, label = failed or not within period). Even a simple decision-tree model might reveal that age >30 combined with insulation=XLPE (old generation) is a high-risk cluster. This gives quick wins in identifying suspects.
3. Use Existing Load Data: Without adding sensors, use substation SCADA and AMI meter data that the utility already collects. Many utilities have hourly load data per feeder and interval meter data per customer. By correlating those, the solution can estimate loading on each cable section (especially for radial systems). Even simpler, if detailed allocation is hard, just use feeder peak load as a percentage of feeder capacity – cables on heavily loaded feeders get a higher stress score. Off-the-shelf, utilities have this from their PI historian or similar. Incorporating it requires some data mapping but no new hardware.
4. Apply Industry Models and Benchmarks: In absence of utility-specific condition tests, the solution can incorporate industry failure models (like IEEE or CIGRÉ guides on cable failure rates by age) to predict remaining life. It can also use known problem lists – e.g., if the utility can list "we have 50 miles of unjacketed PILC cable from 1950s" – industry experience might say those have high failure probability due to corrosion of lead. The software could flag all such segments as red until proven otherwise.
5. Quick GIS-Based Visualization: Using the imported data, the solution can very quickly produce a map of the service territory highlighting the oldest or most failure-prone cable segments. This addresses one of the biggest challenges: just knowing where all the old cables are. Many utilities have this info but not visualized intuitively. A map-based risk heat map can be generated in weeks by combining age and past outages. For example, circuits could be color-coded by "cable health index" derived from age+failures. This immediate visualization often provides insights (e.g., one neighborhood might stand out as all red = needs attention).
6. Prioritization & "Low-Hanging Fruit": With these analyses, the utility can take initial actions even before deploying new sensors. For instance, they might decide to schedule proactive replacements for the worst 5 segments identified (maybe ones that already failed twice). Or, if capital is constrained, they might try cable rejuvenation (fluid injection) on high-risk aged cables to extend life ~10 years at lower cost. The analytical solution can support this by showing how rejuvenation would defer X failures (Siemens' claim of deferring capex by 10 years suggests using such mitigations guided by analytics⁶³).
7. Iterative Improvement with New Data: Once the system is live with basic data, the utility can start feeding new sources as they become available. For example:
   • After a season of smart meter outage data, the system might get more granular fault location info.
   • If the utility does a round of offline tests on sample cables (say tan-delta tests on 20 cables this year), those results can be input and the model refined (and similar cables without tests inferred about).
   • Over time, deploy a few pilot sensors (like install Sentient UM3+ on 2 worst feeders). The data from those pilots can be used to calibrate the model. For instance, the sensors might catch a bunch of momentary faults indicating incipient issues – then the AI knows that pattern and can look for similar patterns in feeder SCADA waveforms (if any).
   • In essence, you start with data at hand, get actionable output in a matter of months, and then continuously enrich the model with better data (sensors, test results, etc.) to improve accuracy.

9.1 Rapid Deployment Feasibility

The good news is, most utilities have the core data needed to start:

• Location / attributes of cables (from GIS)
• Failure history (from reliability reports)
• Some load info (from SCADA / meters)

All these are "off-the-shelf" internal data. The implementation mainly involves data extraction, cleansing, and loading into the analytics platform. Modern utility analytics tools often come with adapters for common systems (GIS, OMS, etc.), making integration faster. For example, if using a cloud-based APM software, you might upload CSV exports of these datasets or connect via API. Within a few weeks the data model can be populated and initial risk scores generated.

Siemens explicitly notes their approach uses "already available data – using advanced analytics and ML"⁶³ to reduce failures by up to 10x. This implies that without any new fancy sensor, significant insights can be gained simply by crunching existing data in smarter ways than before.

To summarize: rapid deployment = leverage existing data + proven algorithms. Do the sophisticated sensor integrations in parallel as a longer-term enhancement. This phased approach gets the utility quick value and builds confidence (which can justify the cost of adding sensors where needed).

Day 1

Utility drops GIS + OMS data into the analytics platform.

Day 2 - 3

Auto-mapper aligns utility schema; analyst validates.

Day 4 - 5

9 Physics AI™ modules run; AHI, RUL, soil & flood scored.

Day 6 - 7

Asset Mgmt, T&D, CIO view per-segment AHI in 3D.

Figure 9.1 · A 7-day onboarding pattern from data drop to live 3D twin. Reference implementation: Orbyfy Earthflow Underground. See it live: earthflow.orbyfy.com/earthflow-underground

Next, we articulate the value proposition and pitch statements that can be used to explain the benefits of implementing such an underground cable analytics solution, followed by an ROI estimation.

11 · Earthflow Underground — Reference Architecture

Chapters 1–9 describe what an underground distribution cable analytics solution should look like in principle. This chapter describes one such solution in production today: Orbyfy Earthflow Underground. It is presented here as a reference architecture — one concrete instantiation of the abstract principles in this paper — not as the only possible design.

10.1 Two-Layer Module Stack: 9 + 15

Earthflow Underground combines 9 cable-specific Physics AI™ modules with a 15-module platform engine. The two layers run together — the Underground modules consume outputs from the platform modules (e.g. soil chemistry, flood probability, lightning density), so the cable-health scoring is grounded in the same authoritative environmental datasets used across Orbyfy's broader product line.

Table 10.1 · The 9 Underground-specific Physics AI™ modules. Every output is per-cable-segment with a provenance JSON pointing to source datasets.

Table 10.2 · The 15 platform Physics AI™ modules underneath. The Underground stack inherits all 70+ authoritative datasets from the platform layer.

10.2 The 7-Day Onboarding Schema Auto-Mapper

The most important architectural choice in Earthflow Underground is the Schema Auto-Mapper. Most utility analytics deployments require multi-month integration projects to align the vendor's data model with the utility's GIS, OMS, SCADA, and AMI schemas. Earthflow's Schema Auto-Mapper detects field-name patterns automatically (Esri ArcFM, Schneider Conduit Manager, generic geodatabases) and produces a candidate mapping in hours, not months. An analyst then validates the mapping in Days 2–3 of the 7-day onboarding, after which the Physics AI™ modules can run.

10.3 Cirra — The Agentic Layer

Cirra is the natural-language orchestration layer that sits on top of all 24 modules (9 Underground + 15 platform). A T&D Asset Management VP can ask Cirra: "Which cables are most likely to fail in storm season?" — and Cirra composes a query across the AHI Composite, Flood Hydrology, and Manhole Risk modules, returns a ranked list of segments with citations back to source datasets (SSURGO, FEMA NFHL, NLDN), and writes a defensible model card explaining the reasoning. This converts a multi-week analyst engagement into a 30-second answer.

10.4 Citeable Provenance JSON — The FERC/PUC Defensibility Story

Every numeric output Earthflow produces carries a provenance JSON block citing the authoritative source dataset, the equation or method used, and the data vintage. When a utility's CIO or Regulatory Affairs team takes an Earthflow-generated cable-replacement recommendation to a rate-case prudency review, the model card is auditable down to the row of the input dataset. This is the foundational design choice that distinguishes Earthflow from black-box ML solutions: regulators cannot defend a proprietary score, but they can defend a citeable physics result.

10.5 Living 3D Twin

Figure 10.1 · Earthflow3D · Living Twin Per-segment AHI is rendered in 3D over the network. Asset Mgmt VPs, T&D Engineers, and OMS Operators each see a tailored view derived from the same underlying model.  ·  See it live →

The visualization layer is a Mapbox-rendered 3D twin tied to the GIS network. Per-segment AHI is color-graded, manhole IoT sensors show real-time gas / temperature status, and clicking a segment opens the full data record including the provenance JSON. The same twin produces tailored views for Asset Mgmt, T&D Engineers, OMS Operators, CFO, CIO, and Reliability Engineers — one network, many decision-ready surfaces.

10.6 Zero New Sensors for v1

Earthflow Underground does not require any new sensors for the v1 deployment. The 7-day twin ships from the utility's existing GIS / OMS / SCADA / AMI fused with 70+ authoritative environmental datasets. Sensor integration (Sentient UM3+, Con Edison-style manhole monitors, Omicron PD systems) is a Phase 2 enhancement that the platform is designed to ingest as those data sources come online.

12 · Case Study — Reference Deployment Pattern

This chapter outlines the deployment pattern for an Earthflow Underground reference engagement. Specific customer numbers are anonymized; the timeline and outcomes are representative of the production beta currently running.

11.1 Deployment Profile

• Service territory: Suburban underground distribution in a coastal Florida service area
• Cable miles: ~6 cable-miles of MV underground
• Customers served on segment: ~12,000 residential and small commercial
• Sensor channels: 187 (mix of intelligent FCIs, AMI dark-sector signals, and manhole IoT)
• Soil profile: Coastal sand-clay, high water table, elevated chloride content (high corrosion risk for older PILC)
• Flood exposure: Mixed AE-zone (FEMA NFHL Layer 28) and X-zone segments

11.2 Onboarding Timeline (Actual)

Week 1 - 4

Onboard + initial AHI scoring across every segment.

Week 5 - 8

Back-test AHI ranking against 5-yr OMS fault history.

Week 9 - 12

Optional: validate against PD + manhole IoT pilot.

Week 13 - 16

Demo Day · signed ROI memo · executive briefing.

11.3 Outcomes (Reference Targets)

• Time to live 3D twin: 7 days from data drop
• Top-50 worst-cable precision (back-test target): ≥70% against 5-yr OMS fault history
• Worst-segment identification: <5% of cable miles account for >40% of historical fault minutes — flagged for proactive intervention
• Deferrable capex identified: $5–15M per 500–1,000-mile pilot territory
• Typical payback: <2 years on the analytics layer alone

11.4 Fleet Health Dashboard — the DPU Prudency View

Figure 11.1 · Fleet Health · DPU / Rate-Case View Portfolio AHI rollup, PILC exposure, material mix, age distribution, and ESMP-aligned benchmarks. The dashboard a Regulatory Affairs lead takes to a state PUC prudency review.  ·  See it live →

Each cell of the dashboard ties back to a provenance JSON that points to the authoritative dataset used. A state PUC asking "how did you decide to defer this $3M cable-replacement project?" gets a citeable answer instead of a black-box score.

13 · Value Proposition and Benefits

Implementing an end-to-end digital analytics solution for underground distribution cables offers a compelling value proposition for electric utilities. Below are key customer-facing pitch statements and benefits:

12.1 Improve Reliability and Reduce Outages

"Proactively prevent cable failures before they happen." An AI-driven analytics solution can identify high-risk cables and allow utilities to fix or replace them before an outage occurs, directly reducing customer interruptions. For instance, predictive analytics using line sensor data have been shown to cut feeder outage minutes by over 20% by preempting faults⁷⁹. Fewer outages mean higher SAIDI/SAIFI performance, improved regulatory compliance, and happier customers. In day-to-day operation, early warning of developing cable issues can avoid emergency failures that cause lengthy unplanned outages.

12.2 Faster Restoration When Failures Do Occur

"Find and fix underground faults in a fraction of the time." By using smart sensors and AI to pinpoint fault locations, utilities can dramatically speed up repair times. This leads to shorter outage durations (lower CAIDI). Some utilities have achieved 50–65% reductions in fault location patrol time with intelligent line monitoring⁵⁸. Customers experience quicker restoration, and utility crews avoid lengthy "blind" digs to find faults. Enhanced visibility translates to higher customer satisfaction and less overtime labor cost.

12.3 Cost Savings in Operations and Maintenance

"Do more with less — target your resources efficiently." An underground analytics solution helps utilities avoid unnecessary maintenance and focus efforts where needed. By tracking the true condition of cables, utilities can move from time-based to condition-based maintenance, saving money. One large utility reported an 11% O&M cost reduction after deploying line sensors and analytics, through more efficient patrols and fewer reactive repairs⁵⁸. Additionally, underground lines typically have lower routine maintenance costs already; analytics further reduce costs by preventing the catastrophic failures that are most expensive. Another study found undergrounding physically eliminates up to 75–80% of maintenance costs²⁸ – our solution helps capture some of those savings even without full undergrounding, by eliminating emergency fixes and optimizing maintenance schedules.

12.4 Extend Asset Life and Defer Capital Expenditures

"Get 10 more years out of your cables safely." With accurate health diagnostics, utilities can safely extend the life of cables that are in decent condition, while only replacing those truly near end-of-life. Siemens' analytics, for example, promise to defer capital replacement by up to 10+ years for cables in the program⁶³. This means millions in capital spending can be postponed or reallocated, improving the utility's capital efficiency. Shareholders and regulators appreciate when expensive cable replacement programs are driven by data – ensuring you're not replacing cables too early (wasting money) or too late (after failure). As a result, the total cost of ownership of underground assets goes down.

12.5 Enhanced Safety and Risk Mitigation

"Prevent manhole explosions and safety incidents." By monitoring conditions like manhole gas and stray voltage, the solution can warn of dangerous situations before they injure employees or the public. Utilities can avoid incidents like electrified manhole covers or fires that pose liability and reputation risk. For example, Con Edison's IoT sensors prevented manhole explosions by alerting crews to intervene in time⁵⁵. Ensuring a safer system protects workers (fewer hazardous "run-to-failure" events in confined spaces) and the community. This also ties into regulatory goals for safety.

12.6 Wildfire and Storm Resilience

"Hardening the grid against disasters." In wildfire-prone regions, underground cables analytics help ensure the buried lines (often installed to replace overhead for wildfire safety) are maintained and won't fail when needed most. By preventing outages, you also reduce ignition risk from failed equipment. In hurricane regions, knowing which underground laterals are likely to fail allows reinforcement before storm season, keeping the lights on when overhead lines might be down. This boosts overall resilience and emergency preparedness.

12.7 Data-Driven Planning and Investment

"Make smarter investment decisions with AI insights." The solution provides utility executives and planners with hard data on asset condition and risk. This means you can justify reliability investments to regulators with concrete evidence (e.g., "We predict 5 cable failures next year on these specific segments – replacing them now will avoid $X in outage costs."). It also helps evaluate the benefit of undergrounding projects – by quantifying how much reliability improvement you get. The transparency and analytics improve regulatory support for proposed budgets, as decisions are backed by data, aligning with the trend of utilities adopting Asset Performance Management for better stewardship⁷³⁷¹.

12.8 Customer Satisfaction and Regulatory Incentives

"Fewer outages and faster restoration means happier customers." Many utilities operate under performance-based rates or scorecards where reliability indices and customer satisfaction scores directly affect earnings or penalties. By implementing this solution, a utility can improve those metrics and potentially earn performance incentives (or avoid penalties). It also provides a positive story to communicate: "We are using cutting-edge AI and sensors to improve reliability," which can bolster customer trust and brand image as an innovative, reliable service provider.

12.9 Ease of Integration and Rapid Benefits

"Leverage what you have — quick to deploy, quick to value." Emphasize that the solution builds on the utility's existing data (GIS, outage info, etc.) and can deliver actionable insights within months, not years. This is not a massive IT overhaul but a modular analytics layer that can start small (on a few circuits) and scale. Quick wins (like identifying a handful of imminent failures and averting them) can pay for the project in the first year. The modular nature also means lower upfront cost compared to infrastructure upgrades.

Each of these points can be part of a pitch to stakeholders (whether internal management, regulators, or customers in communications). The combined value is a smarter underground network that is reliable, safe, and optimized cost-wise.

In quantitative terms, we next estimate the ROI (return on investment) of such a solution, translating some of these benefits into dollar values per feeder or per mile.

14 · Estimated ROI of an Underground Cable Analytics Solution

Investing in an underground cable analytics platform yields returns in several forms: avoided outage costs, reduced labor/O&M, and deferred capital. We estimate the ROI by examining these savings on a per-feeder or per-mile basis and comparing to typical solution costs:

13.1 Avoided Outage Costs (Value of Reliability Improvement)

Every cable failure prevented has a monetary value in terms of avoided customer outage impact and emergency repair cost. According to studies, the economic cost of power outages is very high for certain customers – up to $110,000 per outage for a medium-large commercial business in worst cases²⁸. Even average values are significant (small businesses might lose a few thousand per outage, residential inconvenience is often valued in surveys as well). On a per-feeder basis:

• Suppose a feeder serves 1,000 customers and normally experiences 2 underground cable failures per year, each causing a 4-hour outage for 500 customers (by sectionalizing half the feeder). That's 500 × 4 = 2,000 customer-hours of interruption per event, or 4,000 cust-hrs/year. If we value reliability at (conservatively) $20 per customer-hour (a mix of residential and commercial impact), that's $80,000 economic impact per year from those outages. If analytics can prevent 50% of those outages (say 1 failure avoided), that's $40,000/year of societal value saved for that feeder. For the utility, avoiding an outage also avoids potential regulatory penalties or lost revenue (outages = no energy sold for that duration).
• Utilities also assign customer value metrics like CMI (Customer Minutes Interrupted) improvement. Sentient's data suggests 20% improvement in CMI on instrumented feeders⁷⁹. If a feeder had 1,000,000 customer-minutes of interruption annually (just an example), a 20% reduction saves 200,000 cust-min. At, say, $3 per customer-minute for large industrials or much less for residential, but weighted, perhaps $1–2 per cust-minute on average, that's $200k–$400k value in avoided downtime for that feeder. These numbers are illustrative; actual value depends on customer mix and outage frequency, but clearly non-trivial.

From the utility perspective, regulators may quantify reliability value. For example, Dominion's analysis valued each hour of outage at $35,458 of GDP impact per customer-hour (spread across economy)¹⁷. They got $1.76 benefit per $1 cost in their undergrounding program partly via that metric¹⁷. An analytics solution isn't as dramatic as physically undergrounding, but if it can eliminate some outages, it similarly creates outsized economic value relative to its cost (which is relatively low, mainly software and sensors).

13.2 Reduced Reactive Repair and O&M Costs

When an underground cable fails unexpectedly, the utility incurs significant direct costs:

• Emergency repair crew labor: Overtime for skilled cable splicers, often at night or weekends if it fails then. A single splice repair might take a crew 8–16 hours including travel, setup, locating and fixing the fault.
• Excavation and restoration: If not in an accessible conduit, digging up a street or backyard can cost thousands, plus later repaving.
• Equipment/material: New cable section, splices, etc. Not huge per incident (maybe $1k–$5k materials), but adds up.
• Other costs: Sometimes a failed cable can cause collateral damage (fire damage to nearby equipment, or require environmental cleanup if any insulating fluid/oil is involved for older cables).

Anecdotally, an unplanned underground failure repair can cost anywhere from $5,000 (simple dig and splice) to $50,000+ (if extensive civil work or overnight contractor needed). Let's estimate $20,000 average cost per underground fault for the utility (labor, equipment, restoration). If the analytics prevents, say, 5 failures per year across the system, that's $100,000 saved right there in direct costs. Larger utilities with dozens of failures could save much more proportionally.

In addition, line sensors reduce patrol costs: One utility cited 65% patrol time reduction⁵⁸. Crews that used to spend, e.g., 6 hours searching for a fault might now spend 2 hours. If each hour of crew time is $200 (multiple crew members and vehicles), saving 4 hours per fault is $800. Across 50 faults a year, that's $40k saved. And that's just labor; faster restoration also means fewer switching operations, fewer customer calls (reducing call center load), etc.

Utilities also spend on preventive maintenance that could be optimized. For example, some do routine manhole inspections every few years. If IoT sensors reduce how often physical inspections are needed (say extend interval from annually to every 3 years), that's fewer truck rolls. The solution acts as a "virtual inspector"⁸³, saving potentially hundreds of field visits, which could easily be tens of thousands of dollars annually.

Overall, one can envision an 11% O&M cost reduction due to line sensing and analytics, as was reported by a southeastern utility⁵⁸. If a utility's distribution O&M budget is $50 million, 11% is $5.5M/year saved – a huge ROI considering an analytics solution might cost on the order of a few hundred thousand to a million to implement (depending on scale and hardware).

Even focusing on underground-specific O&M: if a utility spends $10M/year on underground repairs and maintenance, reducing that by 10% (through smarter targeting and fewer failures) saves $1M/year.

13.3 Capital Deferral and Life Extension

This is more abstract but very potent financially. Replacing underground cable is expensive capital work – often $200k to $1M per mile depending on context⁸⁵⁸⁶. If an analytics solution allows a utility to defer replacing 10 miles of cable by 5 years (because it determined those cables are still okay), that defers say $5 million of spending. The net present value saved can be a big portion of the solution's cost. Siemens noted up to 10-year deferral of capex per cable⁶³. Even deferring one major network rebuild project yields ROI because the utility can use that capital elsewhere or avoid financing costs. On the flip side, the solution also prevents wasted capex on the wrong cables (ensuring you don't replace a healthy cable while missing a hidden failing one). So capital is spent more efficiently – in effect increasing ROI of your capital by targeting it better.

13.4 ROI Calculation Example

Consider a utility implementing the solution on 50 feeders (maybe a mid-size utility). Suppose:

• Cost of solution: $1 million upfront (software, integration, some sensors) + $200k/year software/SaaS and maintenance. Over 5 years, total cost ≈ $2 million.
• Benefits per year:
  • 10 avoided cable failures (out of perhaps 40 that would have happened): direct saving ~$200k (10 × $20k each).
  • Improved reliability: 10 fewer failures * 1000 customers * 2 hr avg outage = 20,000 cust-hrs avoided. Value maybe $50 per cust-hr for commercial-heavy feeders (or use the Dominion $35 per customer-hour for GDP which would be $700M – too high for ROI internal calc, but anyway). Let's conservatively say $200k economic value per year or maybe $50k of utility financial incentive (like avoiding performance penalties).
  • O&M labor savings: fewer patrol hours and targeted maintenance, save 500 labor hours ~$50k.
  • Deferred capex: perhaps avoid replacing a few cables prematurely, saving $500k in a given year of the plan.
  • Total quantifiable benefit per year ~ $200k + $50k + $50k + $100k (if we amortize the deferred capex benefit per year) = $400k (very rough).

Over 5 years that's $2M benefit vs $2M cost – break-even. But these numbers were conservative. If the solution is more effective (like the 11% O&M reduction on $50M budget = $5M/year saving noted earlier), the ROI becomes multiple times the investment. In many cases, the payback period can be under 2 years. For example, one underground cable failure causing a big downtown outage can cost enormous amounts in claims or regulatory fines; if the solution prevents even one major event (say a hospital outage or a downtown network explosion), it may pay for itself right there.

Another way to present ROI is per feeder or per mile:

• Cost per feeder to instrument and include in analytics might be, say, $20k (covering sensors and share of software). If that feeder avoids one outage of 1,000 customers for 4 hours, using customer interruption cost of ~$10 per customer-hour (blended), that outage avoidance is worth $40k economically. That's a 2:1 benefit-cost for that one incident. If it avoids multiple outages and improves restoration times routinely, the ratio grows.
• Cost per mile of cable in the program – maybe $5k/mile in analytics costs. If failures are say 0.1 per mile-year normally, preventing half of those (0.05 failures avoided) each costing $20k repair and $20k customer value = $40k *0.05 = $2k benefit per mile-year. Over a 5-year period = $10k benefit per mile, vs $5k cost = 2x ROI. And that's without considering big outlier events or the intangible benefits of safety and reputation.

13.5 Real-world Competitive Context

It's also worth noting that competitors (Hitachi, GE, Schneider) presumably would have case studies demonstrating ROI of their solutions. For instance, GE's APM software highlights general benefits like 10–40% reduction in reactive maintenance⁸⁷. If we take 20% reduction in reactive work for cable maintenance, that's directly cost saving. Schneider's pitch for EcoStruxure Asset Advisor often cites preventing failures that cause costly downtime – e.g. preventing one fire or avoiding one unplanned outage on critical equipment far outweighs the solution cost. And Hitachi's Lumada APM emphasizes optimizing maintenance spend and extending life⁸⁸⁸⁹, which from a CFO perspective is a clear financial win (maximizing return on assets).

13.6 ROI Beyond Dollars — Strategic ROI

Some returns are harder to monetize but important:

• Regulatory goodwill: showing regulators that the utility is innovating to improve reliability can support rate cases. Some jurisdictions might even allow adding these investments to rate base or provide incentive rate of return for reliability improvements.
• Customer trust: fewer outages and proactive communication about preventing issues (e.g., "we identified a failing cable and replaced it with minimal downtime") improve customer sentiment, which has long-term value for utilities, especially as they seek EV adoption, demand response participation, etc. Satisfied customers are more cooperative customers.

In conclusion, while exact ROI will vary, an underground cable analytics solution can reasonably deliver a return on investment of several times its cost over a few years. It does so by slashing unplanned outage costs, trimming maintenance fat, and squeezing more life out of existing assets. This high ROI potential is driving all major grid technology vendors to offer solutions in this space, which we examine next in terms of competitive offerings.

15 · Where the Incumbents Stop — Siemens, Hitachi Energy, GE Vernova, Schneider Electric

Four incumbent vendors. One shared gap. Several major players offer solutions for grid asset analytics and monitoring, each with their own emphasis. Below we compare Siemens, Hitachi Energy (formerly ABB Power Grids), GE (now GE Vernova), and Schneider Electric in the context of underground distribution cable analytics and management — then map where each one stops short of a complete, 7-day-deployable, multi-peril-fusion, FERC/PUC-defensible solution.

14.1 Siemens

Siemens leverages its deep power-engineering expertise and digital portfolio under initiatives like Siemens Xcelerator and Advanta services. Specifically for underground cables, Siemens offers a Cable Analytics solution that focuses on data-driven failure prediction. According to Siemens, their approach uses advanced data analytics and machine learning on readily available utility data (cable attributes, loading, etc.) to identify at-risk cables and optimize maintenance⁶³. They claim this can "reduce cable failures by up to a factor of 10" and extend asset life by deferring replacement by 10+ years⁶³. This bold claim indicates Siemens has developed predictive models (likely from their experience with utilities worldwide) that dramatically improve reliability when applied.

In implementation, Siemens can provide end-to-end support: installing sensors (if needed), integrating data, and running analytics. For example, Siemens can deploy SICAM Fault Indicators (EFIs) for underground networks – these are devices that detect earth faults in MV cable networks and interface with Siemens automation⁹⁰. Siemens also has offerings in distribution automation (SCADA, FLISR systems) which complement cable analytics by automating isolation of faults. They tie this into a unified ADMS or asset management system. Additionally, Siemens has products for thermal monitoring of power systems and digital twin software that can simulate cable aging under different load and environmental conditions.

A differentiator for Siemens is its strong focus on integration of OT data with AI. Their solutions often involve both a software component (analytics algorithms, dashboards) and an advisory service via Siemens Advanta to help interpret results and implement changes. They also emphasize achieving policy or strategic goals: e.g., their advocacy for "50% underground by 2040" came with suggestions to use "underground-line-sensing technology married with artificial intelligence" to ensure quality installation and operation⁹¹. This indicates Siemens not only supplies the tools but also thought leadership on how to transform the grid with digital tech.

14.2 Hitachi Energy (formerly ABB)

Hitachi Energy offers solutions under its Lumada Asset Performance Management (APM) and enterprise software portfolio. Lumada APM is a next-generation platform that Hitachi has been evolving (the latest release announced Oct 2023)⁷³. It provides a unified view of asset health, reliability, and optimized investment planning⁹²⁷³. For underground cables, Hitachi's approach would integrate data into Lumada APM's health and prognostics models. They highlight that Lumada APM uses AI-driven prognostics and statistical models to predict failures and advise proactive maintenance⁷¹. The platform can give prescriptive recommendations – e.g., it might tell the utility "Cable X has 80% probability of failure in next 12 months – schedule replacement" – and allow scenario analysis (what if we replace now vs later)⁷¹⁹³.

Hitachi Energy also has a strong pedigree in grid monitoring hardware from its ABB heritage. They produce a range of sensors and distribution automation gear: for example, ABB's GridShield reclosers, RIO switches, and fault indicators that can be applied to laterals, including underground circuits. ABB had developed underground fault detection systems (like the ABB AFI fault indicator). We can infer Hitachi Energy integrates these with their Network Manager ADMS and Lumada analytics. They recently have promoted "digital substation" concepts and fiber optic sensing in some contexts⁹⁴⁹⁵. ABB also has experience in underground HV cables monitoring for HVDC projects; that expertise can translate to distribution scale (technologies like distributed fiber sensing for temperature or detecting sheath currents).

While we don't see a product marketed solely as "underground cable analytics" from Hitachi, their APM solution is generic across assets and can certainly encompass cables. Hitachi's differentiation might be its combination of APM with field service management (FSM) and enterprise asset management (EAM) software (they offer a full suite: Lumada APM, Lumada EAM, etc.⁹⁶). This means their solution can tie cable analytics directly into work management – when a cable is flagged, an inspection work order can be auto-created in the EAM system. Also, Hitachi's focus on sustainability and risk (they mention reducing waste, optimizing energy, safety in APM benefits⁹⁷⁹⁸) aligns well with prolonging cable life and avoiding disruptive failures, which can be pitched as both an economic and environmental win (less frequent cable manufacturing and trenching is a sustainability gain).

14.3 GE Vernova (General Electric)

GE's offerings in this area fall under its Asset Performance Management (APM) software and Grid Analytics. GE's APM (which includes the platform formerly known as Meridium, now part of GE Digital) is a leading platform recognized by analysts (Verdantix, etc.)⁹⁹. It is typically used in generation and transmission, but GE has been extending it to distribution grid assets as well. GE Vernova emphasizes that its APM suite "enables increased reliability, decreased maintenance costs, [and] lowered risk" across energy assets¹⁰⁰. These are exactly the aims for underground cable management.

GE's competitive edge might come from its holistic GridOS software portfolio, which includes ADMS (Advanced Distribution Management System) and specialized analytics like Distributed Energy Resource management and advanced protection. If a utility already uses GE's ADMS, integrating an asset analytics module for cables could be smoother. GE also historically has offered services like reliability studies and survival analyses through entities like GE Consulting or partnerships with EPRI. They can bring statistical models to the table. For sensors, GE does not make small distribution sensors like SEL or others, but they do have Intelligent Breakers and Relays that record disturbances (for example, GE Multilin relays at substations can capture oscillography of faults). GE's SmartSignal predictive analytics (from the industrial side) could also be adapted for electrical anomalies – it's an AI platform used to predict failures in rotating machinery, but GE might apply similar approach to find patterns in grid data.

One should note GE's APM has been used by utilities like Ameren to monitor T&D assets¹⁰¹. The approach was to gather condition data from multiple sources (dissolved gas from transformers, etc.) and unify it. For cables, GE would do similarly: unify whatever condition data (test, inspection, thermal) into their APM health scoring. GE's marketing highlights outcomes like 20% reduction in reactive maintenance at some customers¹⁰² and various percent improvements in availability and cost reduction⁸⁷, which would apply to the distribution context as well.

In summary, GE's solution is part of a broader electric utility software suite (GridOS APM) and they would likely compete on being an integrated platform (from planning to operations) and on their analytics pedigree. They might not have as specific a cable-focused product as Siemens' consulting offering, but they cover the need through APM customization.

14.4 Schneider Electric

Schneider Electric's strengths lie in distribution network GIS and operational technology, plus their evolving digital services (EcoStruxure). Schneider is unique in owning ArcFM (a widely used GIS for utilities) and related tools like Conduit Manager³⁸. This means Schneider can manage the detailed data of underground networks extremely well – mapping ducts, cables, manholes, etc.¹⁰³⁴⁶. They emphasize efficient data handling of the physical infrastructure, which is a prerequisite for any analytics. For example, Conduit Manager allows a utility to accurately maintain where each cable is located and its occupancy of ducts¹⁰⁴ – this is vital for a map-based analytics solution.

On the analytics side, Schneider's EcoStruxure Asset Advisor / Advisor services provide predictive analytics and remote monitoring for electrical distribution assets. Their marketing for digital services states it uses AI and machine learning, remote monitoring, predictive analytics to improve uptime¹⁰⁵¹⁰⁶. They mention concepts like Asset Health Index and Maintenance Index to track condition and wear¹⁰⁷¹⁰⁸. Specifically, Schneider highlights monitoring partial discharge in medium-voltage switchgear as part of their offering¹⁰⁹, which is closely related to cables (the same technology can be applied to cable insulation). It stands to reason that Schneider's partial discharge monitoring could extend to cable joints and terminations as well, since they produce PD sensors (often in partnership with companies like IPEC or EA Technology for continuous monitoring).

Schneider is also heavily involved in power quality and energy monitoring (for facility power systems), which gives them experience with sensors measuring electrical parameters. While Schneider might not have a branded "underground cable analytics" product, they integrate asset analytics into their EcoStruxure ADMS and services. They have case studies in condition-based maintenance where their digital services prevented failures. For example, Schneider launched EcoStruxure Asset Advisor initially focusing on critical assets like UPS and motors¹¹⁰, but they have expanded it to electrical distribution gear – including transformers (EcoStruxure Transformer Expert)¹¹¹, etc. They can easily incorporate cable condition into that umbrella.

One competitive advantage Schneider might tout is its end-to-end solution from GIS design to field execution. A utility could use Schneider's ArcFM to design cable replacements, use EcoStruxure ADMS to operate the network, and Asset Advisor to monitor cable condition – a one-stop shop. Schneider's ArcFM also now has web and mobile components, meaning field crews using Schneider's mobile GIS could see the analytics output in the field (e.g., a crew gets a mobile map highlighting a risky cable they are near).

Moreover, Schneider has a strong service arm for remote monitoring – their Connected Services Hub where experts monitor customer assets 24/7¹¹²¹¹³. They could extend that to utility cable networks, effectively offering a service where Schneider helps watch the cable fleet and notify the utility of issues (though utilities might prefer to keep that in-house).

In summary, Schneider competes by combining data management (GIS) with analytics and services. They emphasize empowering maintenance teams with AI-driven insights and improving uptime via an Asset Health Index approach¹⁰⁷. For underground cables, Schneider would leverage its deep knowledge of how to document those assets (ArcFM Conduit Manager) and pair it with their predictive analytics engines and IoT devices (they manufacture things like temperature and PD sensors for gear which could apply).

14.5 The Shared Gap — What None of the Incumbents Deliver

Each of these companies is thus positioned slightly differently: Siemens and GE as more software/analytics-centric (with Siemens also strong in field devices), Hitachi as a blend of software and device with a focus on unified APM, and Schneider as the GIS+OT incumbent adding AI services.

From a utility's perspective, a competitive evaluation might look at ease of integration (e.g., if they already use Schneider GIS, Schneider's solution could plug in nicely; if they use GE ADMS, GE's might), the maturity of the analytics (Siemens' bold claims vs. need to see proof), and overall cost/licensing models (some may be cloud subscription like GE's, others more custom project like Siemens Advanta consulting).

Ultimately, the trend is that all major vendors recognize underground cable analytics as a crucial need and have built or are building capabilities to address it. This competitive push is good for utilities, as it means solutions are becoming more readily available and feature-rich – helping turn the vision of a smarter, more resilient underground grid into reality.

But the shared gap remains. Across all four incumbent platforms, none deliver all six of the following in a single product:

Table 14.1 · Capability matrix · The combination only Earthflow Underground offers in a single product. ✓ = full capability · ~ = partial / requires extensive customization · ✗ = not in product.

This is the gap Earthflow Underground was built to close. The next chapter describes the moat.

16 · Where Physics AI™ + Agentic AI Begin — Earthflow's Differentiator

The market has incumbent point-tools. None of them are wrong — they are simply incomplete for the specific job of producing a defensible, deployable, multi-peril Underground risk twin. This chapter describes the architectural design choices that distinguish Earthflow Underground from the incumbents and from any foundation-model lab that might enter the space.

15.1 The Moat Is Not the LLM

15.2 Citeable Provenance — The Regulator-Defensible Moat

The deepest design choice in Earthflow Underground is that every numeric output emitted by the platform traces to a citeable authoritative dataset. The provenance JSON for an Asset Health Index of 4.2 on cable segment XYZ cites: SSURGO horizon-by-horizon chemistry for the corrosion sub-score, NLDN regional flash density for the lightning sub-score, FEMA NFHL Layer 28 for the flood-exposure sub-score, the utility's own OMS for the historical-failure sub-score, IEEE C57 / C37 standard cable-aging curves for the age sub-score. Nothing is a proprietary black-box weight.

Why this matters: when a utility's Regulatory Affairs lead walks into a state Public Utility Commission rate-case prudency review and says "we deferred replacing these 47 miles of cable and accelerated replacement on these other 31 miles," the PUC's first question is "on what basis?" — and the auditable answer must trace back to a public dataset and a published equation, not to a vendor's proprietary score. Earthflow's provenance JSON makes that the default. None of the incumbent solutions surveyed in Chapter 15 ship provenance at the equation level.

15.3 Physics-Rooted Modules — Not Statistical Regression

Each of the 9 Underground modules described in Chapter 11 is grounded in a published equation or empirically-validated method. The Cox proportional-hazards survival model (Module 2 · RUL Survival) has been the standard cable-survival framework in the IEEE distribution literature for two decades. The soil corrosion kinetics (Module 5) draw from established sheath-loss equations against SSURGO chemistry. The FEMA NFHL Layer 28 flood-zone overlay (Module 6) is the authoritative US flood-hazard data source.

The contrast with a pure-ML approach: a deep-learning model trained on 10 years of utility outage records can certainly find statistical patterns, but it will struggle to give an auditable reason for any specific prediction. Earthflow's physics-rooted modules tell you both what and why.

15.4 Cirra — The Agentic Layer That Actually Earns Its Keep

The market is awash in "AI assistants" bolted onto enterprise software that effectively wrap search-over-documentation. Cirra is different: it is an orchestration layer over 24 Physics AI™ modules. When a T&D Asset Mgmt VP asks Cirra "which cables are most likely to fail in storm season?", Cirra:

1. Decomposes the question into the relevant module queries (AHI Composite · Flood Hydrology · Manhole Risk · Wind·Load · Hail·WRF)
2. Executes those queries against the live data layer for the utility's network
3. Ranks the resulting segments by combined risk score
4. Composes a written response with inline citations back to each contributing module and source dataset
5. Saves the conversation as a defensible model card for the utility's records

The end result: a 30-second answer to a question that would otherwise be a 2-week analyst engagement, with the audit trail attached.

15.5 The Reinsurance Adjacency — Same Engine, Different Buyer

Figure 15.1 · Earthflow Underwriter · Reinsurance vertical The same Physics AI™ engine powers underground cable risk for utilities and treaty-pricing for reinsurance underwriters. The physics that says "this cable is at risk" is the same physics that says "this exposure has this expected annual loss." One platform, two adjacent markets.

This is the strategic depth of the architecture: the 70+-dataset provenance graph and the multi-peril fusion engine that power Underground also power Earthflow Underwriter (reinsurance) and Earthflow Detect (EPC bid evaluation). The same physics that flags a cable segment as high-risk drives a treaty premium calculation on the reinsurance side. One engine, multiple adjacent verticals.

15.6 What's Hard — Honestly

This paper would be poorer if it did not name what's hard about the Earthflow approach:

• Schema alignment with utility GIS / OMS is non-trivial. The Schema Auto-Mapper handles common cases (Esri ArcFM, Schneider Conduit Manager, generic geodatabases), but edge cases — legacy custom schemas, partial GIS coverage, OMS exports lacking event-level metadata — still need analyst hours in Days 2–3. The 7-day commit holds for ~80% of utilities surveyed; the long tail requires custom mapping.
• Sensor data integration is Phase 2, not v1. The 7-day twin ships from existing utility data plus environmental layers; Sentient UM3+, manhole IoT, and Omicron PD streams are added when the utility chooses to deploy them. The platform is designed to ingest those data sources, but they are not in the v1 deployment.
• Provenance is only as defensible as the source datasets. SSURGO, FEMA NFHL, NLDN, USGS 3DEP, IEEE C57 / C37 are all authoritative. But cable-internal data (test results, partial-discharge history) often resides in unstructured PDFs in utility document-management systems — getting that into the provenance graph is an ongoing effort, not a solved problem.

The honest summary: Earthflow Underground is not the first analytics platform for underground distribution cable. The incumbents have point tools that solve pieces of the problem well. What Earthflow does is fuse those pieces into a 7-day-deployable, multi-peril, physics-rooted, FERC/PUC-defensible whole — and put an agentic-AI Q&A layer on top. The moat is the data graph plus the tradecraft, not the model architecture. See it live →

17 · Conclusion — A Citeable, Physics-Rooted Path Forward

Underground distribution cable analytics is emerging as a critical component of modern grid management. As utilities in the U.S., Canada, and worldwide bury more of their networks for resilience, the ability to digitally monitor and optimize these invisible assets will differentiate industry leaders. An end-to-end solution — combining asset data, smart sensors, and AI-driven analytics with intuitive map-based tools — offers a path to dramatically improved reliability (fewer and shorter outages), cost savings, and extended asset life.

The technology discussed in this report shows that such solutions are no longer theoretical; they are being deployed with measurable success (20%+ reliability gains, 10× failure reductions in trials, etc.). By investing in underground cable analytics, utilities can turn the challenges of underground systems (difficulty of access and detection) into an opportunity — using data to shine a light underground, so to speak. The result is a safer, more reliable, and more efficient power delivery system, better equipped to meet the demands of the 21st century.

16.1 What This Paper Asked, and What It Found

This paper began with a comprehensive question — what does it take to digitally manage North America's 1.1+ million miles of underground distribution cable? It surveyed:

• The scale and trajectory of undergrounding (Chapters 1–2)
• The current organizational, budgetary, and visibility challenges (Chapters 3–4)
• The state of data collection, KPIs, and sensor hardware available today (Chapters 5–7)
• The architecture an AI/ML solution would need and how to deploy it rapidly (Chapters 8–9)
• A reference implementation — Orbyfy Earthflow Underground (Chapters 10–11)
• The business case — value proposition and ROI math (Chapters 12–13)
• The competitive landscape and the differentiator gap (Chapters 14–15)

The case studies and statistics cited make a strong argument that the ROI is attractive and the need is pressing, especially as electrification and resilience goals push us toward greater undergrounding. The incumbent vendors — Siemens, Hitachi Energy, GE Vernova, Schneider Electric — have produced credible point tools, but none of them ship a 7-day-deployable, multi-peril-fusion, FERC/PUC-defensible solution as a single product. The market still has a real gap.

16.2 The Path Forward

The integration of end-to-end digital analytics in the underground distribution space is not just a nice-to-have — it is rapidly becoming a best practice for leading utilities aiming to provide top-tier reliability and customer service in an era of aging infrastructure and climate challenges. Adopting these solutions will ensure that as we build more underground lines, we also build the intelligence to manage them effectively, closing the loop on a truly smart and robust grid.

Orbyfy Earthflow Underground is one credible answer to the gap. Citeable provenance instead of black-box scoring. 7 days instead of 7 months. 9 cable-specific Physics AI™ modules on top of a 15-module platform engine. FERC and PUC defensible by design. One living twin, six decision-ready surfaces — for the T&D Asset Management VP, the OMS Operator, the CFO, the CIO, the Reliability Engineer, and the field crews who actually go fix the cables.

16.3 Call to Action

18 · FAQ for Utilities

18.1 What data is already in place before we ever talk to the utility?

Earthflow Underground starts with a data foundation we have already gathered, cleaned, and kept current. Nothing on this list requires anything from the utility — it is all in place on the day the conversation begins. Each dataset feeds one or more of the nine Physics AI™ models, listed here alongside the answer it produces.

Every input carries provenance metadata. For any score we report, we can show the utility exactly which dataset contributed and where we used an industry default instead of the utility's own history.

18.2 What does the utility need to provide — and how do we handle it?

We work with whatever the utility can provide, in whatever shape it is in. Underground distribution data is rarely clean — that is a structural feature of the asset class (see Chapter 6), not a utility-specific failing — and we never ask a utility to scrub its records before sending them to us. The table below lists the inputs that compound value fastest, the ones that sharpen results when they happen to exist, and the ones we explicitly do not want the utility spending effort on.

Bottom line: six of the nine Physics AI™ models go live from the cable map and outage log alone — even when those records carry the typical underground-data gaps. The daily-load readings and smart-meter data light up the remaining three.

How we handle the data itself

Underground distribution data is heterogeneous, partial, and often decades stale. Our architecture assumes that on day one, not as an exception to be managed:

• Schema Auto-Mapper. Detects field-name patterns across the common cable-mapping, outage-log, and historian schemas — Esri ArcFM, Schneider Conduit Manager, generic geodatabases, and in-house variants. Produces a candidate mapping in hours, not months. We target 80% schema coverage in the first 48 hours; an analyst-led validation pass closes the remaining 20%.
• No common data model required. We do not ask utilities to conform to one canonical schema. Earthflow runs on a semantic knowledge graph — heterogeneous source systems plug in through the Auto-Mapper and are reconciled in graph space, not by force-fitting fields into ours.
• Built for bad data. When utility-specific records are missing or stale, we impute from international failure curves (IEEE and CIGRÉ) and the data foundation in 18.1, and surface the gap as an explicit confidence interval — not hidden behind a precise-looking number.
• Provenance per output. Every score traces back to the inputs that produced it — which dataset contributed, where industry defaults filled a gap, and how stale each input was. Engineers, regulators, and rate-case reviewers can audit the basis of any specific recommendation.
• We do the unstructured extraction. Partial-discharge test reports, manhole inspection logs, and other artifacts stored as PDFs, scans, or paper attachments are parsed by us. Utilities never need to digitize first.

Seven days from data drop to a live risk twin. That is the engineering choice that makes this possible.

18.3 Do you need historical data to identify alarms or similar events?

To start, no. Six of the nine Physics AI™ models run on day one using the data foundation above plus the utility's cable map. Where the utility's own failure history is missing, we fall back to internationally cited industry failure curves (IEEE and CIGRÉ standards). The utility receives a ranked, citeable list of its worst-5%-of-segments within seven days of the data drop — even with zero labeled failures on file.

For a defensible back-test and tighter accuracy, yes. Our primary success measurement — getting at least 70 of every 100 “worst-cable” predictions correct against the utility's own historical fault locations — only becomes possible once five years of outage history is in hand. With that history, the Remaining Useful Life model calibrates to the utility's own vintages and failure patterns. Without it, results are still produced; they simply carry wider confidence intervals, surfaced explicitly in every output rather than hidden behind a precise-looking number.

In short: historical data is what proves the worst-5% list is right. It is not what makes the list possible.

For the deployment timeline behind these answers, see Chapter 11 — Earthflow Underground Reference Architecture and Chapter 12 — Case Study: Reference Deployment Pattern.

About Orbyfy Earthflow · About the Authors

About Orbyfy Earthflow

Orbyfy is a Boston-based Climatetech company building Earthflow — a Physics AI™ + Agentic AI living digital twin for critical infrastructure. The platform fuses 70+ authoritative environmental datasets with utility, EPC, and reinsurance data through a 15-module Physics AI™ engine and an agentic AI orchestration layer (Cirra). Two product verticals are in production today: Earthflow Underground (underground electric distribution cable risk) and Earthflow Underwriter (reinsurance treaty pricing). Orbyfy was founded in 2025, is a member of Greentown Labs in Somerville, MA, won Edison Awards Gold 2026, and is part of NVIDIA Inception and the Black & Veatch IgniteX accelerator cohort. Production runs on Google Cloud Platform with NVIDIA NIM-compatible inference paths. The company is partnered with EPRI (AI for Power Challenge 2026 submission, Category 1 · Grid Asset Management) and is assembling a multi-utility Demonstration Partner Consortium across the Northeast.

About the Authors

Hara Brar

Founder · CEO · CTO · Orbyfy

20+ years at the intersection of power grid and AI — Siemens, Alstom Power, GE, Hatch. IDC-ranked digital solutions architect for utility infrastructure. Hands-on developer behind the Earthflow geospatial core and the Underground Physics AI™ engine. Leads product, engineering, and customer engagements. Boston, MA.

Greg Elliott

Founder · Chief Physics AI™ Officer · Orbyfy

20+ years across fin-tech, energy & utilities, and applied Physics AI™ research — Bloomberg, Thomson Reuters, Siemens, Cinchy, BMO, SAS Institute. Architects the Physics AI™ stack — 15 physics-informed modules fusing authoritative environmental datasets with utility telemetry into AHI, RUL, and risk forecasts.

Glossary of Acronyms

References & Data Sources

Supporting data and examples have been drawn from industry reports, utility case studies, vendor literature, DOE and regulatory reports, and product information from Siemens, Hitachi Energy, GE Vernova, and Schneider Electric, among others. These illustrate the current state and benefits of underground cable analytics in practice.