Orbyfy Earthflow™
Live demo ↗ White paper ↗ Underground Demo 1 / 7
Earthflow Underground

A Physics AI living digital twin for primary medium-voltage underground distribution.

Every segment scored. Every score auditable back to the inputs behind it — measured or imputed. Earthflow Underground is a solution for Underground Cable Health.

The Solution — value delivered
1
Where to underground

Which corridors buy the most risk reduction per mile of spend.

2
Where to test next

Rank segments for partial discharge and tan δ, instead of testing on the calendar.

3
Condition

A health score on every segment, with remaining useful life and a confidence band.

4
Capacity bottlenecks

Where headroom runs out — before the new load arrives to find it.

Live twin Earthflow Underground twin of a distribution network, with cable routes, vaults and layer controls
Earthflow Underground, running — synthetic beta feeder
What we offer
Data as a service · scores delivered by API or export Agentic AI · ask the network in plain language Integration-ready · sits alongside your GIS, OMS, SCADA and AMI No new sensors · no rip-and-replace
The Problem

~1.1 million miles of underground distribution cable — and utilities cannot see any of it.

GIS only A distribution network as a utility holds it today: bare cable routes on satellite imagery, with every condition, sensor and environmental layer switched off
Your network today — routes, and nothing else
Condition unknown

Replaced on age, not evidence — and an underground fault costs 3–4× an overhead one to restore.

Capacity unknown

Headroom judged on nameplate ratings — while electrification and new load keep arriving.

Testing means an outage

VLF, tan δ and sheath integrity all need the segment de-energized. Most utilities test a fraction and run the rest to failure.

+5,000–10,000 miles a year are going underground. The blind spot in the power grid is growing.

That spend is regulator-approved and rate-recovered — FPL's Storm Secure Underground Program, Dominion's SUP, PG&E's plan. Every dollar has to survive a prudency review, where age alone is weak evidence.
The Hypothesis

A cable does not fail because of itself. It fails on electrical and environmental conditions — together.

ChemicalSoil that eats a sheath — chloride, sulfate, acidity, resistivity.
HydrologicStanding water in a flooded duct; groundwater cycling year on year.
ThermalSoil that will not carry the heat away.
ElectricalSurge exposure over the route; load transferred when a parallel feeder trips.

Environmental data is ambient and system-wide — never resolved to the segment. That is the gap.

⚡ BURIED DISTRIBUTION CABLE ◆ splice joints water table pH 5.2Cl⁻ H₂OpH 6.8 ⚡ ELECTRICAL 🌍 ENVIRONMENTAL ⚛ FUSION
The physics of ageing

Cable ageing is textbook physics — Neher–McGrath thermal rating, Arrhenius thermal ageing, inverse-power electrical stress, corrosion kinetics, Weibull survival. An electrical · thermal · chemical · mechanical problem, acting in combination — exactly what Physics AI is built for.

PILC · paper-insulated, lead-covered
  • Corrosive soil attacks the lead sheath — moisture reaches the paper, and dielectric strength collapses.
XLPE · cross-linked polyethylene
  • Water treeing — moisture under electrical stress grows trees through the insulation until breakdown.
The Environmental Intelligence Layer

Earthflow powers granular geographic site intelligence — environmental perils on infrastructure. Acre by acre.

Running today Heat-map analysis view: a site divided into a grid of cells, with a long list of environmental analysis layers available down the side
Heat-map analysis on the solar vertical — grids down to ~1.6-acre cells, imagery at 1–10 m per pixel

70+ data sources · 200+ KPIs per location · 1,100+ data points per lat/long · 1,300 sites · 1M+ acres analysed

19+11
Physics AI modules — 30 in total
19 across the platform, plus 11 built for underground cable · continental US coverage · answers in minutes, not months
What we fuse together
Utility data · four extracts
  • Asset registry · route & vintage
  • Outage history · 5–10 yr, the ground truth
  • Feeder loading · ≥1 yr, 15-minute
  • Meter intervals · de-identified load shape
Our environmental layer · fields per site
  • Soil chemistry · 26
  • Soil moisture · 44
  • Flood exposure · 49
  • Precipitation · 36
  • Ground movement · 24
  • Surface erosion · 60
Minimum viable — asset registry + 5 yr outages + 1 yr loading. Supply a subset to start; the floor is geometry and attributes.

Schema Auto-Mapper aligns whatever fields you have — 80% mapped in 48 h, an analyst validates the rest. No common data model required, and no staff time on data prep.

Fused per segment — it resolves to your cable route, not a regional average.

What You Receive

The same four questions — now answered on every feeder segment.

1
Where to underground

Corridors ranked by risk reduction per mile of spend.

Capital planning output: optimal annual capex against reactive cost, outage minutes avoided, customers protected
Ranked corridor list
2
Where to test next

The worst-segment list your partial-discharge and tan δ crews work from.

Ranked segment table with per-segment action: replace, defer or maintain, each with a confidence figure
Ranked test queue
3
Condition

Vintage, your outage history, soil corrosivity and flood exposure — with remaining useful life.

12 34 5 healthy end of life
Health Index 1–5 · what you replace
4
Capacity bottlenecks

A thermal rating from soil properties and burial method, against your measured loading.

measured load headroom thermal rating
Capacity Index, % headroom · what you connect

Produced by eleven physics-grounded modules — AHI Composite · RUL Survival · PD Analytics · Soil Corrosion · Flood Hydrology · Manhole Risk · Dual-Basis Thermal Rating · Loading & Headroom · Fault Triangulation · AMI Dark-Sector · Capex Deferral.

Every score carries a value, a confidence interval and its provenance — each input tagged measured or imputed, so a number can be defended rather than just quoted.

Up to 10 years of deferred capital, and ~10% off underground O&M (modelled)

The Twin, Running

A living twin of your network — every segment scored, every score auditable.

Rendered on a synthetic beta feeder. Click any segment for its index, its confidence band, and the inputs behind it — measured or imputed.

Open the live explorer →
Let's Get Started

Start with four extracts. We do the rest — no new sensors.

Utility data
  • The four extracts from slide 4 · registry, outages, loading, meter intervals
  • Read-only, one-way, no operational-technology link
Earthflow Environmental Intelligence Layer
  • Soil · chemistry & corrosivity
  • Water · flood & groundwater
  • Moisture, terrain · surge exposure
  • 100% national coverage, resolved to the segment
Earthflow Physics AI — live in week one
Eight underground modules running in your first week, eleven in total on the 30-module platform · injected equations you can audit, never black-box
Health Index 1–5 · remaining useful life Capacity Index · % headroom to new load
Keep every asset · no rip-and-replace Messy data welcome · we map it Read-only, one-way · never touches operations Delivered your way · API or export

Hara Brar · Founder & CEO · hara@orbyfy.com · www.orbyfy.com · Orbyfy, Inc. · Boston, MA

Edison Awards 2026 Gold EPRI AI for Power Finalist Black & Veatch IgniteX Siemens Next47 Start-up School NVIDIA Inception Boston Business Journal IDC Leader · Utility Solutions Techstars EnergyTech Greentown Labs
Click or press → to advance