Spatial Intelligence Engine · zooid.net Layer

Location is not a property.
It's a dependency graph.

The same sensor reading means something completely different depending on where it was taken, what's upwind, what terrain intervenes, which jurisdiction governs it, and what happened at neighboring sensors sixty seconds ago. geog.ai resolves that dependency graph — in real time, for every reading, on every device.

zooid.net
sensor ingestion
geog.ai
spatial meaning
m8n.app
autonomous decision
50 ZOOID Verticals · 6 Propagation Engines · Continuous Calibration Loop · Real-time Jurisdiction Mapping

What geog.ai is — and is not

What Stripe did to payment APIs, geog.ai does to spatial intelligence — turns something that required specialist tooling into a clean API call that any application can make.

geog.ai IS geog.ai is NOT
A real-time spatial reasoning engine A GIS tool for analysts
A physics-aware decision substrate A mapping application
An API layer that turns location into context A tile server or visualization platform
A propagation and simulation runtime A replacement for QGIS, ArcGIS, or ESRI
A calibration loop that improves with sensor data A static geodata warehouse

ArcGIS answers: What is where?
geog.ai answers: Given where and current conditions — what happens next, to whom, and who is responsible?
That is a categorically different product.

Every IoT platform stores what and when.
Most stop there.

The full decision loop requires a third axis — without it, you have a data logging system, not a decision engine.

// A gas sensor reads 50 ppm H2S. That alone tells you almost nothing.

// Resolve the WHERE — now you have a decision:
{
  "wind_vector": { "speed_ms": 8.2, "bearing_deg": 312 },
  "source": { "asset": "wellpad_07", "distance_m": 195 },
  "terrain": { "ridge_height_m": 15, "slope_deg": 3.1 },
  "upwind_sensors": [{ "reading_ppm": 0 }, { "reading_ppm": 0 }],
  "jurisdiction": { "authority": "TCEQ", "limit_ppm": 10, "notify_within_hr": 24 },
  "personnel_zones_at_risk_in_12min": ["Zone C", "Zone D"]
}

The raw sensor value was one field. The rest is spatial context — resolved, enriched, and current. That's what geog.ai provides.

The Four Where-Types

Every spatial system that fails does so because it treats location as a single coordinate field. geog.ai treats location as having four distinct semantics, each requiring different storage, indexing, and reasoning.

Where-Type 1

Observation Geometry

Where the measurement was physically taken. Includes lat/lon/altitude for outdoor sensors, and building/floor/zone/room for indoor. This is the starting point — a spatiotemporal fact that changes when devices move.

Where-Type 2

Asset / Site Geometry

Where the thing being monitored exists. A well pad is a polygon. A pipeline is a polyline. A watershed is a network. A factory fence is a boundary. This geometry is stable but versioned — it changes when the physical asset changes.

Where-Type 3

Field-of-Influence Geometry

The modeled footprint of a phenomenon — a plume, a flood inundation extent, an RF coverage map, a noise contour, a thermal diffusion field. This geometry is computed, not stored. It changes as conditions change.

Where-Type 4

Governance Geometry

Jurisdictional boundaries, permit areas, exclusion zones, regulatory districts, easements. Location determines which rules apply, which authority is notified, which permit threshold governs. Versioned and auditable — legally consequential.

Six engines that model how phenomena move through space

Most IoT platforms stop at spatial indexing and context enrichment. geog.ai goes further: physics-based simulations of how things propagate through terrain, atmosphere, and infrastructure.

C1 · Atmospheric

Atmospheric Dispersion

Gaussian plume (Tier 1, <1s), AERMOD regulatory grade (Tier 2, 1–30s), and Lagrangian particle tracking for non-steady-state scenarios. Calibrated per-site against actual sensor readings.

POST /simulate/plume

Try the live plume demo
C2 · RF / Wireless

RF Propagation

ITM/Longley-Rice terrain model for outdoor links. HATA/COST-231 for urban deployments. Coverage probability polygons, link budget analysis, mesh viability matrix, and node placement optimizer.

POST /simulate/rf_coverage

POST /rf/mesh/viability

POST /rf/optimize/placement

Try the live RF coverage demo
C3 · Hydrology

Flood & Hydrological

DEM-based watershed delineation, SCS Curve Number rainfall-runoff, kinematic wave channel routing, and 2D flood inundation mapping. EPA SWMM integration for urban stormwater networks.

POST /simulate/flood

Try the live flood inundation demo
C4 · Acoustic

Noise Propagation

ISO 9613-2 standard octave-band outdoor sound propagation. Accounts for geometric spreading, atmospheric absorption, ground effect, barrier insertion loss, and vegetation attenuation. Produces isodB contour polygons.

POST /simulate/noise
C5 · Thermal

Thermal Diffusion

Heat propagation from industrial sources, fire events, or HVAC failures. Time-to-temperature-threshold at monitored locations, hot zone polygons for personnel exclusion, and cooling cascade prediction.

POST /simulate/thermal
D · Graph

Spatial Graph Engine

Models sensor networks as graphs — nodes (devices/assets), edges (distance, radio, fluid, causal influence). Enables influence spread queries, evacuation routing, anomaly clustering, and mesh failure analysis.

GET /graph/influence

GET /graph/evacuate

The calibration loop — the feature that compounds

GIS platforms don't have sensors. IoT platforms don't have physics models. geog.ai has both — and runs a feedback loop that neither competitor can replicate.

Simulate plume → predict concentrations at sensor locations
                            ↓
           Compare against actual sensor readings
                            ↓
                 Compute residuals (predicted - observed)
                            ↓
      Update site-specific model parameters (stability correction,
      terrain factor, effective emission height)
                            ↓
               Next simulation is more accurate
1
Site-calibrated models — earned, not bought

After 30 days, plume predictions are site-calibrated. After 90 days, they're calibrated to seasonal meteorological patterns. After a year, they're calibrated to facility operating modes. A competitor deploying the same software on the same site starts from scratch.

2
Jurisdiction database depth

Regulatory boundaries, permit IDs, notification thresholds, cross-boundary relationships — assembled from hundreds of public sources, normalized, versioned, and continuously maintained. Expensive to replicate.

3
Sensor network topology graph

The spatial graph of deployed devices, their connectivity, their historical behavior patterns, their terrain relationships — a proprietary asset that grows with every deployment.

4
Compounding gap

A new competitor on day one has generic models. geog.ai on year two has site-specific models calibrated against millions of actual sensor readings. The gap widens every day the system runs.

geog.ai × m8n — spatial layer meets decision layer

geog.ai is the spatial layer. m8n is the decision layer. Neither is complete without the other. Every m8n native function becomes location-aware through geog.ai.

m8n.app

Decision engine, workspace orchestration, blueprint execution, outcome learning

↓ calls

geog.ai

Spatial context resolution, propagation modeling, RF optimization, jurisdiction mapping, calibration loop

↓ receives from

zooid.net

Edge agents, sensor ingestion, mesh networking, Scout/Hive/Swarm/Colony tiers, IoT verticals

Blueprint + geog.ai

Spatial Triggers

Blueprints can fire on spatial conditions — "plume crosses permit boundary," "asset enters exclusion zone," "RF coverage drops below minimum in zone." Jurisdiction-aware notifications route to the correct authority based on where the event occurred.

Pattern Analysis + geog.ai

Terrain-Normalized Anomalies

A reading that looks anomalous is normal if the sensor is in a valley during a temperature inversion. Pattern Analysis accuracy improves when geog.ai resolves whether a reading is spatially expected or genuinely anomalous.

Bootstrap + geog.ai

Location-Driven Configuration

When Bootstrap receives a deployment location, geog.ai resolves jurisdiction (sets thresholds), terrain type (selects propagation model), local climatology (sets dispersion parameters), and neighboring deployments (identifies calibration references).

Optimization + geog.ai

Sensor Placement Intelligence

Bootstrap's initial deployment recommendation includes geog.ai's sensor placement optimization — given terrain, prevailing wind, and asset layout, where should the first sensors go for maximum coverage and minimum redundancy?

The verticals where WHERE changes everything.

Every zooid.net vertical has at least one spatial dependency. These are the verticals where geog.ai's spatial reasoning is the critical path — where the sensor reading means nothing without the plume direction, watershed position, regulatory boundary, or terrain between source and receiver.

🌿 Environmental & Safety
V01 Gas & Air Quality Monitoring

Primary: Atmospheric dispersion — plume simulation and wind-source triangulation. Three sensors reading zero upwind + one reading 50 ppm downwind = source location, plume direction, emission rate estimate, and twelve-minute risk forecast for personnel zones.
Secondary: Jurisdiction mapping for TCEQ/EPA notification thresholds. Personnel zone proximity from indoor spatial graph.

POST /simulate/plume GET /jurisdiction GET /nearby
V02 Water Quality & Hydrology

Primary: Watershed delineation and inundation mapping — every water sensor's position resolved to its upstream contributing area and downstream flood zone. Contamination plume tracked through the hydrological network.
Secondary: Upstream sensor graph for causal attribution. NPDES permit boundary enforcement via governance geometry.

POST /simulate/flood GET /jurisdiction GET /context
V06 Weather & Microclimate

Primary: Terrain-adjusted spatial interpolation and kriging — microclimate readings from a valley floor mean something fundamentally different than readings from a ridge top. geog.ai resolves terrain context for every met station to produce spatially meaningful nowcasts.
Secondary: Fire weather index enrichment. Atmospheric stability class resolution for plume models that depend on this vertical's wind data.

GET /context GET /nearby POST /simulate/plume
V50 Wildfire Detection & Prescribed Burn Management

Primary: Fire spread polygon and ember transport zone — combining fire source, terrain, and wind vectors to model propagation in minutes. Ignition risk enrichment from lightning strike coordinates and fire weather index. Prescribed burn containment geofence and escape-route check.
Secondary: CAL FIRE/USFS district boundary and notification authority. Upstream weather station graph for fire weather index computation.

POST /simulate/plume GET /within GET /jurisdiction GET /nearby
🏗️ Infrastructure & Assets
V04 Seismic & Geotechnical Monitoring

Primary: Vibration spatial correlation and settlement mapping — seismic signals propagate through soil in ways that depend on geology, depth, and intervening structure. geog.ai resolves which sensors are geologically co-located vs. separated by a fault plane or soil type boundary.
Secondary: Building footprint and infrastructure polygon context. Slope stability zone geometry for landslide-risk enrichment.

GET /context GET /nearby GET /within
V23 Pipeline & Linear Asset Monitoring

Primary: Spatial correlation of pressure delta anomalies along a polyline asset to locate leak position. A pressure drop between sensors narrows the leak to a segment; geog.ai resolves that segment geometry, its terrain, and nearest access point.
Secondary: PHMSA permit boundary and right-of-way polygon. Nearest water body and settlement for consequence-of-failure enrichment.

GET /context GET /jurisdiction GET /within GET /nearby
V36 Power Generation & Grid Infrastructure

Primary: Transmission line coordinate enrichment for ice-load and wind-load per segment using terrain and weather context. Power network graph for outage propagation modeling — which downstream assets are at risk if this transformer fails.
Secondary: FERC territory polygon for jurisdiction and notification authority. Terrain-adjusted wind resource via spatial kriging for renewable integration planning.

GET /context GET /within GET /nearby GET /jurisdiction
🚛 Mobility & Transport
V19 Traffic & Transportation Intelligence

Primary: Route analysis and congestion zone detection — traffic counts are only meaningful when resolved to their road network position, intersection geometry, and upstream/downstream flow relationships. geog.ai resolves the infrastructure graph that makes adaptive signal control spatially coherent.
Secondary: DOT district jurisdiction for incident notification. V2X RSU coverage polygon for autonomous vehicle corridor analysis.

GET /context POST /simulate/rf_coverage GET /jurisdiction
V22 Marine & Port Operations

Primary: Vessel trajectory modeling, berth proximity resolution, and plume dispersion over water for exhaust and fuel spill analysis. Tidal current modeling affects both plume transport and berthing safety margins.
Secondary: Tidal zone jurisdiction for COLREGS and port authority notification. Hull sensor positions resolved against berth and channel geometry for collision-risk enrichment.

GET /trajectory POST /simulate/plume GET /jurisdiction GET /nearby
🌾 Agriculture & Environment
V21 Precision Agriculture & Land Management

Primary: Field zone analysis and soil variability mapping using NDVI + terrain + soil sensor spatial interpolation. Variable rate application zones are output geometries — polygons that prescribe exactly where and how much to apply, computed from spatial heterogeneity in the sensor network.
Secondary: Watershed delineation for irrigation return flow modeling. Property boundary for regulatory compliance on chemical application.

GET /context GET /within POST /simulate/flood GET /jurisdiction
V27 Aquatic & Oceanographic Monitoring

Primary: Water body geometry and current modeling for drift prediction of buoys, gliders, and pollutant plumes. DO and temperature readings from distributed sensors spatially interpolated across the monitoring area to build a continuous field model — not isolated point readings.
Secondary: EEZ and watershed boundary for regulatory trigger evaluation. Effluent plume simulation for pollution source tracing.

POST /simulate/plume GET /context GET /within GET /trajectory
V45 Aquaculture & Fish Farming

Primary: Pen GPS coordinates resolved against tidal current vector and water exchange rate — the same DO reading in a well-flushed pen and a stagnant corner means something completely different. Spatial interpolation of DO and temperature across the pen array identifies which pens need intervention.
Secondary: Aquaculture licensing authority and lice treatment trigger via watershed/EEZ boundary. Effluent plume for downstream impact assessment.

GET /context GET /within GET /nearby POST /simulate/plume

All 50 zooid.net verticals have spatial dependencies resolved by geog.ai.

These are the verticals where the spatial component is the critical path. Every other vertical benefits too — from jurisdiction-aware notifications to sensor placement optimization to indoor-outdoor position unification.

Explore all 50 verticals at zooid.net ↗

Quick reference — primary geog.ai call by vertical

Vertical Primary geog.ai Call Secondary
V01Gas & Air Quality Monitoring
POST /simulate/plume
GET /jurisdiction
GET /nearby
V02Water Quality & Hydrology
POST /simulate/flood
GET /graph/upstream
GET /within (watershed)
V04Seismic & Geotechnical Monitoring
GET /context
GET /nearby (soil zone)
GET /within (slope stability)
V05Noise & Vibration Monitoring
POST /simulate/noise
GET /jurisdiction (noise limits)
V06Weather & Microclimate
GET /context
GET /nearby (terrain-adjusted)
V17Emergency & Disaster Response
POST /simulate/plume (multi-source)
GET /graph/evacuate
GET /jurisdiction (dynamic)
V18Drone Operations & Aerial Intelligence
GET /rf/coverage
GET /trajectory/predict
GET /within (no-fly zones)
V19Traffic & Transportation Intelligence
GET /context
POST /simulate/rf_coverage (V2X RSU)
V21Precision Agriculture & Land Management
GET /coverage (field zones)
GET /context (soil variability)
POST /simulate/flood (watershed)
V22Marine & Port Operations
GET /trajectory
POST /simulate/plume (spill)
GET /jurisdiction (tidal)
V23Pipeline & Linear Asset Monitoring
GET /context (spatial correlation)
GET /jurisdiction (right-of-way)
V27Aquatic & Oceanographic Monitoring
POST /simulate/plume (ocean)
GET /trajectory (buoy drift)
GET /within (EEZ boundary)
V36Power Generation & Grid Infrastructure
GET /context
GET /within (FERC territory)
GET /graph/influence (outage)
V47Border & Perimeter Security
POST /rf/mesh/viability
POST /rf/optimize/placement
GET /rf/gaps
GET /rf/coverage
V50Wildfire Detection & Prescribed Burn Management
POST /simulate/fire
POST /simulate/plume (smoke transport)

All 50 ZOOID verticals ship with geog.ai context resolution on every sensor event. See zooid.net/zooid/#clusters for the full vertical explorer.

Spatial intelligence as clean API calls

Every response includes the result, a confidence score (0.0–1.0), model version, calibration age, and full data provenance.

# Context resolution
GET  /context?device_id=sensor_h2s_023
GET  /context/batch?device_ids=[...]

# Proximity and spatial queries
GET  /nearby?lat=31.96&lon=-99.90&radius_m=500&type=gas_sensor
GET  /within?geometry={polygon}&type=asset
GET  /jurisdiction?lat=31.96&lon=-99.90

# Propagation simulations (async jobs with webhook callback)
POST /simulate/plume     { source, wind, stability_class, duration_min }
POST /simulate/flood     { watershed_id, precip_mm, duration_hr }
POST /simulate/rf_coverage { node_id, frequency_mhz, tx_power_dbm }
POST /simulate/noise     { source_id, octave_bands_db }
POST /simulate/fire      { ignition_point, fuel_model, weather }

# RF-specific (propagation, mesh, placement)
GET  /rf/coverage?node=N&threshold_dbm=-90
GET  /rf/link?from=A&to=B
POST /rf/mesh/viability      { node_ids, failure_scenarios }
POST /rf/optimize/placement  { area, coverage_threshold, existing_nodes }
GET  /rf/gaps?area=polygon&threshold_m=500

# Optimization and calibration
POST /optimize/node_placement  { area, coverage_threshold, existing_nodes }
POST /optimize/sensor_placement { vertical, target_coverage_pct }
GET  /model/calibration?model=plume&site_id=wellpad_07

Pay per call. Subscribe per site.

We sell spatial decisions — not maps, not tiles. Two pricing dimensions: metered API usage for queries and simulations, and site-level calibration subscriptions that compound over time.

Usage Metered
Context API
$0.001 / call

Spatial context resolution for any device. Returns full SpatialState — terrain, jurisdiction, neighbors, wind vector, confidence score.

GET /context per device
GET /nearby proximity queries
GET /jurisdiction at any coordinate
Full data provenance on every response
Confidence score included
Get API Key
Simulation Metered
Propagation Runs
$0.05–5 / run

Physics-based simulation jobs — plume, flood, RF coverage, noise contours, fire spread. Priced by model tier and complexity.

Gaussian plume (Tier 1): $0.05/run
AERMOD regulatory grade (Tier 2): $0.25/run
Flood / watershed simulation: $0.10/run
RF mesh viability matrix: $0.15/run
Node placement optimization: $1–5/run
View Full Rate Card →
Endpoint Type Rate Notes
GET /context Per call $0.001 High volume, low margin — designed for every sensor event
GET /jurisdiction Per call $0.0005 Cached aggressively; jurisdictions change rarely
POST /simulate/plume Per run $0.05–$0.25 Tier 1 (Gaussian) vs Tier 2 (AERMOD regulatory)
POST /simulate/flood Per run $0.10 Includes watershed delineation + inundation mapping
POST /simulate/rf_coverage Per run $0.08 Single node coverage polygon at given RSSI threshold
POST /rf/mesh/viability Per run $0.15 Full mesh connectivity matrix with failure probability
POST /optimize/node_placement Per run $1.00–$5.00 Consultants charge $50K+ for this manually
Site Calibration Per site/mo $200–$500 Compounds in value — models improve every month

Bundle discount: 20% off geog.ai API usage when paired with an active zooid.net Swarm or Colony tier subscription. Site calibration subscriptions are priced by vertical complexity — gas/plume sites start at $200/mo, multi-vertical enterprise sites at $500/mo.

Spatial intelligence as a dependency, not a product

geog.ai is the spatial layer your ZOOID verticals already need. Context resolution, propagation simulation, jurisdiction mapping, and a calibration loop that compounds — available as clean API calls, starting today.

Request API Access Explore zooid.net verticals ↗ See m8n Platform

Part of the zooid.net · geog.ai · m8n.app stack.