THE PLATFORM
Predicterra runs a living simulation of Earth and your infrastructure, helping you anticipate what is coming and manage operations ahead of time.
LIVING EARTH + NETWORK MODEL
WATER
Flood flow and terrain exposure
FIRE
Ignition and spread pressure
HEAT
Stress on assets and operations
TERRAIN
Spatial context down to assets
HOW IT MOVES
05 / ACT
API + workflows
LIVE SOURCES
EARTH SYSTEM MODEL
OPERATIONAL SYSTEMS
THE GROUND SHIFTED
The energy industry is being squeezed harder than at any point in its history, and from more directions at once. The physical world turned unpredictable. The hundred-year flood arrives twice a decade. Fire season opens a month early. The correlations that held for fifty years stopped holding, and the tools built on them never saw it coming. At the same time, the ground rules changed. Regulators demand proof of resilience, not promises. Sustainability and disclosure obligations turn every climate exposure into something you have to measure, report and defend. And the energy transition is loading the grid with more assets, more distributed, more exposed than it was ever designed to carry. Every one of these pressures traces back to the same place: the physical state of the world and how it acts on your network. Model that, and you answer all of them at once. Fail to, and you are managing each with a different tool, none of which agree.
HISTORICAL ENVELOPE / BROKEN ASSUMPTION
PAST RANGE ≠ FUTURE STATE
THE ARCHITECTURE
Predicterra is built as four layers that compound: a structure for the world, a simulation of its physics, an intelligence that learns it, and a coordinate system that pins it to your assets.
LAYER 1 — ONTOLOGY
The world as a graph, not a table
Predicterra holds your network as connected entities and relationships: assets, terrain, hazards and the links between them, not rows in a database. The model knows how a transformer depends on a feeder, and how that feeder crosses a floodplain. So risk moves through it the way it moves through the real network.
graph.entities + relationships
LAYER 2 — SIMULATED PHYSICS
A living digital twin, driven by physics
Over that structure, Predicterra runs the physical forces that drive risk: water and flooding, fire ignition and spread, heat, and soil saturation. It simulates how those forces behave. It does not correlate when they last happened. That is why the twin holds under conditions your history never contained.
physics.force_fields + boundary_conditions
LAYER 3 — AI
Intelligence that learns your network
AI builds, calibrates and tunes the models against how your network actually behaves, closing the loop between what the physics predicts and what your grid does. It adapts to your assets and their real thresholds instead of forcing them into a template. Every new data point makes it sharper.
model.calibration + feedback_loop
LAYER 4 — GEOSPATIAL ENGINE
Pinned down to the asset
Every entity, process and prediction is resolved in space and time, down to the asset. A hazard is not a regional warning. It is a specific line, at a specific place, at a specific hour. Risk becomes a coordinate you act on, not a heat-map you interpret, translate by hand or reconcile across disconnected tools.
geo.coordinates + asset_resolution
THE TECHNOLOGY BEHIND IT
None of this is off the shelf. The ontology, the simulation engine and the geospatial core were engineered from the ground up by Woza to model the physical world. Predicterra is that technology, pointed at energy.
WOZA TECHNOLOGY
INSIDE
Built by Woza
01
Ontology
02
Simulated physics
03
Planetary geospatial core
THE LINE
This is the line, and every tool in the category falls on the wrong side of it. A statistical model asks: given a past like this, what usually happens? A simulation asks the only question that matters now: given the state of the world right now, what happens next? Under a stable climate you could get away with the first. That climate is gone. When conditions move past anything in the record, correlation breaks and simulation holds, because physics does not care whether something has happened before. It only cares what forces are in play. One model memorised the past. The other understands the world. Only one of them is still right when the weather stops repeating itself.
CORRELATION
Memorises the past. Breaks outside the record.
SIMULATION
Understands forces. Predicts what happens next.
WHAT IT NEEDS
Start with the data your teams already maintain. Predicterra connects it to Earth observation and simulation layers, then sharpens as more data comes online.
CLIENT INFRASTRUCTURE LAYER
Start with the data you already maintain
The platform can start incomplete and sharpen as more data connects. Each input improves the ontology, the simulation, the thresholds and the actions your teams receive.
incomplete → connected → calibrated
asset_registry
Asset IDs, types, ownership, criticality, maintenance status and operational metadata.
geo_coordinates
Exact asset locations, routes, corridors, sites, substations, plants, pipelines or facilities.
thresholds
Heat tolerance, flood exposure, wind limits, access constraints and outage sensitivity.
incident_history
Observed failures, interruptions, inspections, claims, near misses and known weak points.
enterprise_systems
SCADA, GIS, EAM, ERP, data lakes and planning systems.
WHAT COMES OUT
OUTPUT 01
Asset-level risk scores
Every hazard resolves to the assets it threatens, with severity and confidence attached.
OUTPUT 02
Event impact forecasts
See what an incoming physical event is likely to do to your network before it arrives.
OUTPUT 03
Cascading dependency maps
Understand what sits downstream of each exposed asset, feeder, site or zone.
OUTPUT 04
Operational actions
Turn model output into inspection, dispatch, planning, compliance or control-room decisions.
OUTPUT 05
API-ready model data
Push risk events, thresholds, assets and time windows directly into your existing systems.
OUTPUT 06
Audit-ready evidence
Keep a clear record of exposure, model reasoning and response for reporting and regulators.
ALREADY BUILT. ALREADY RUNNING.
Predicterra does not arrive as an empty engine waiting for your data. It arrives wired to the world’s observation networks and loaded with models ready to run on your assets on day one.
Wired to the world’s satellites
Live feeds from NASA, ESA, Copernicus and meteorological services, fused into one coherent state of the planet. Sourced, cleaned and held live, so your team never touches a pipeline.
7000 data sources, always on
7000
Model library, ready to run
Models across flood, wildfire, wind, heat, soil and vegetation, built for real infrastructure and ready on day one. Not scoped, not rebuilt per project. Ready.
30+ models, out of the box
30+
Every hazard, one model
Flood, fire, wind, heat, drought and vegetation act together. Predicterra models them over one network, so compound and cascading risk is visible instead of scattered across tools.
6 hazard types, on one model
6
SYSTEM OBJECT
Predicterra does not stop at dashboards. Each prediction becomes a structured object that can move through SCADA, GIS, asset management, planning and enterprise systems without your teams translating it by hand.
risk_event.json
LIVE OUTPUT
01 {
02 asset_id: substation_42,
03 hazard_type: flood,
04 severity: high,
05 confidence: 0.91,
06 time_window: 2026-02-14T06:00Z / 18:00Z,
07 downstream_assets: 184,
08 recommended_action: inspect_and_preposition_crews
09 }
ENGINEERED FOR PRODUCTION
A platform built for the systems, controls and scale that critical infrastructure runs on.
01
Meets your security and compliance controls
Hosted in the EU and validated on Microsoft Azure. Your data stays in European jurisdiction, encrypted and governed to the standards critical infrastructure demands.
02
Runs on any cloud, or your own servers
Deploy on any major cloud or on-premise in your own data center. The platform lives wherever your security and sovereignty policies say it lives.
03
An API that connects to everything
A full API drives risk, models and data straight into the SCADA, GIS, asset and enterprise systems you already run. It connects to your stack. It does not sit beside it.
ALREADY OPERATIONAL
REAL GRID DEPLOYMENT PLACEHOLDER
Operational proof, not a promise
Add the Iberdrola proof here: which part of the network is modeled, which hazards are covered, and what measured result the platform produced. Add real scale here too: assets modeled, geographies covered and data sources integrated.
NOT A PROMISE. ALREADY OPERATIONAL.
Predicterra already runs with real operators — Iberdrola, Naturgy, Moeve. It’s certified on the Microsoft Azure Marketplace, a member of GSGIH and FutuRed, winner of the PERSEO innovation award and the BIND energy challenge. Powered by Woza, and built on NASA and ESA data — physics, not correlation.
READY BEFORE THE NEXT EVENT