The spatial intelligence layer for AI

Teaching AI tounderstand maps

Ryshu is building a spatial reasoning system that enables autonomous agents to model maps, geometry, moving assets, and the relationships between them — in real time.

Read the thesis
3Dgeometric reasoning
<50msstate propagation
graph depth
The spatial gap

AI can see images.
Not space.

Modern AI describes a map as pixels. It cannot reason about geometry, orientation, topology, movement, or the relationships between moving assets — the primitives the real world runs on.

01 / GEOMETRY
Geometry

Volume, orientation, navigability.

02 / TOPOLOGY
Topology

Nested assets and containment graphs.

03 / MOVEMENT
Movement

Kinetic state, routing, constraints.

04 / RISK
Risk

Disruption propagation across the network.

LLMsees a flat image of a terminal
RYSHUmodels the topology, constraints and latency of the port
Core insight

Every moving asset
is a graph of value.

Legacy systems track carriers and cargo as separate rows. Ryshu models them as interconnected entities inside a single spatial graph — so AI can reason about how disruptions propagate through the whole network.

nested asset depth
Object graph // MARS-VOYAGER
SHIP · CARRIER40.71°N · 74.00°W
CONTAINER_ALPHA · UNIT 48
ID_892 · HIGH_VALUE_ELECTRONICS
ID_441 · THERMAL_COMPONENTS
CONTAINER_BETA · UNIT 12
ID_102 · CRITICAL_INFRASTRUCTURE
CONTAINER_GAMMA · UNIT 24
ID_553 · PERISHABLE_GOODS · 4°C
STATE: PROPAGATINGLIVE
Process

Model.Reason.Decide.

Use cases

Built for
operations.

From contested airspace to global port networks, Ryshu deploys wherever autonomous systems must reason about the real world.

01 / DEFENCE & AUTONOMY

Mission-aware autonomy.

Operational coordination and spatial reasoning under dynamic constraints.

02 / LOGISTICS NETWORKS

Cargo-level intelligence.

Real-time disruption propagation across multi-modal supply chains.

03 / MARITIME & AVIATION

Dynamic pathing.

Routing across ports, restricted airspace and constrained infrastructure.

04 / DRONE OPERATIONS

3D fleet coordination.

Path optimisation and autonomous coordination in high-density environments.

05 / WAREHOUSING & ROBOTICS

Machine-readable movement.

Spatial intelligence for AMRs and industrial robotics.

06 / SMART INFRASTRUCTURE

Digital-twin reasoning.

Spatial modelling for cities, utilities and critical national systems.

Platform
Geometric topology sphere with orbital paths

Infrastructure
for spatial AI.

Deploy Ryshu as the spatial reasoning layer beneath your fleet, your autonomy stack, or your operational AI.

4surfaces

One spatial reasoning engine — exposed through four deployment surfaces to match how your stack consumes it.

<50msSpatial state propagation
99.99%Reasoning availability
waitlist
Enterprise PlatformOperational intelligence UI
waitlist
Spatial APIReasoning primitives
waitlist
Autonomy LayerEmbedded SDK for fleets
waitlist
Edge RuntimeOn-vessel and on-vehicle
IN DEVELOPMENTPre-launch · design targets

Spatial reasoning,
in real time.

Real-time
Spatial decision engine
designed for continuous inference
target rolling SLA
Reasoning availability
99.99%
p99 design target across regions
State propagation
< 50ms
Operational surfaces
Differentiator

Not another
dashboard.

Ryshu replaces human-driven monitoring with machine-readable operational intelligence. We don't show where things are — we tell the AI what they're doing.

Capability
Legacy
Ryshu
Asset visibility
Static point tracking
Dynamic relationship graph
Operational reasoning
Human-in-the-loop
Autonomous spatial logic
Object model
Flat record structure
Multi-layer nested ontology
Map archetype
Visual UI for humans
Machine-readable geometry
Decision velocity
Minutes to hours
Real-time propagation
Thesis

Language models gave machines an understanding of text. Ryshu gives them an understanding of the physical world itself.

Build the spatial
intelligence layer.

Partner with us to deploy the operational brain for the physical world.

Talk to engineering

Limited partnerships · defence, maritime, autonomy