What We Work On

Capability
areas.

Autonomy isn't one thing — it's several fields working together, from the sensors that take in the world to the software that decides what to do. These are the areas we focus on. For a plainer, general explanation of the technology, see our Technology page.

Domain 01

Autonomy Core

The decision-making part of a system — how it plans a task, adapts as things change, and controls its own movement with limited human input. This is the core of what we work on.

PlanningWorking out routes and actions, and adjusting them as the situation changes.
NavigationKnowing where it is and finding its way, including when signals are unreliable.
Bounded behaviourOperating inside clear limits, with actions that can be reviewed afterwards.

AutonomyThe Loop

LoopSense → decide → act, repeated

ControlKeeping the system stable and safe

ConnectivityDesigned to work with or without a link

OversightA person can supervise and step in

Domain 02

Perception & Sensing

A system can only act on what it can sense. Perception is about combining cameras and sensors into a reliable picture of the surroundings — and knowing how much to trust it.

Sensor fusionCombining different sensors into one clearer picture than any single one gives.
Detect & trackRecognising and following objects of interest in the scene.
Knowing the limitsUnderstanding when the data is uncertain, and flagging it.

PerceptionSensing

Domain 03

Edge Intelligence

Running the intelligence on the device itself, rather than in the cloud. This keeps a system working without a constant network link and keeps its data on board.

On-deviceModels designed to run on small, low-power hardware.
Improves with testingLearning from test data to make the next version better — the idea behind our name.
Works offlineDoesn't depend on a live connection to do its job.

Edge AIOn-Device

RunsOn the device, in real time

HardwareSmall, low-power computers

DataKept on the device

ImprovesLearns from test data, human-reviewed

Domain 04

Resilient Systems

Bringing hardware, software and safety together into something dependable. The aim is systems that keep working sensibly when things go wrong, and that meet the standards defence needs.

SecureGood security practices built in from the start.
Fail-safeSensible, safe behaviour if a part fails.
ModularDesigned to fit together with other systems, not lock you in.

RESILIENCEHARDENED

Where We Apply It

Shared foundations,
different bodies.

The same underlying work — autonomy, perception, systems engineering — applies across the systems we're developing. See the Systems page for more.

§ 05 / SYSTEMS

S-01

Unmanned Aerial

Autonomous aerial platforms (UAVs) for observation and defence tasks.

S-02

Humanoid

Human-like robotic platforms — a research focus on mobility and control.

S-03

Tactical Miniature

Small-form autonomous systems for constrained, close-range settings.

Common Questions

Straight answers.

Our main focus is the autonomy — the software and engineering that make a system autonomous — and we're applying it to autonomous bodies (unmanned aerial, humanoid and tactical). We're at an early, development stage.

Doing the computing on board (rather than in the cloud) means a system can keep working without a constant network connection, and its data can stay on the device. It's a sensible default for defence use.

Our view is simple: machines can carry out tasks, but people set the rules and stay responsible. We design systems so a person can supervise, and so decisions can be reviewed afterwards.

We're not publishing specifications or claiming a finished product at this stage. If you're a partner, supplier or program and want to talk, please get in touch.

Work With Us

Want to work
with us?

If autonomy is relevant to what you do, we'd be glad to talk.

Capabilityareas.