POMERION

Architecture

Between intelligence
and execution.

Pomerion is not an autonomy stack and not a robot OS. It is the authorization layer that sits between them — so every physical action passes through an explicit decision.
The stack

Four layers. One boundary.

L1

Autonomy Layer

Decides what it wants to do

  • VLA Models
  • Planning
  • Navigation
  • Perception
  • Agents
▾

L2

Pomerion

Decides whether it can

  • Machine Identity
  • Model Identity
  • Attestation
  • Policy Engine
  • Authorization Runtime
  • Deployment Control
  • Audit
▾

L3

Machine Runtime

Executes authorized commands

  • ROS / ROS 2
  • Robot Controller
  • Vehicle Controller
  • Flight Controller
  • Edge Runtime
▾

L4

Physical World

Consequences

  • Motors
  • Actuators
  • Movement
  • Manipulation
  • Vehicles
  • Machines
Primitives

Four distinct questions.

01

Identity establishes who or what is acting.

Each machine holds a hardware-rooted, cryptographically verifiable identity, distinct from the operators and cloud workloads around it.

02

Attestation establishes what is running.

Before authority is granted, the machine proves its hardware, firmware, software and model state.

03

Policy determines what is permitted.

Declarative rules bind machines and models to capabilities, resources, zones and conditions.

04

Authorization determines whether the physical action happens.

Every request is evaluated at the boundary, before commands reach the controller.

Request lifecycle

The life of an authorization decision.

  1. step 01

    Intent issued by autonomy stack

  2. step 02

    Identity verified

  3. step 03

    Attestation checked

  4. step 04

    Policy evaluated in context

  5. step 05

    Decision returned: allow / deny

  6. step 06

    Command released or withheld

  7. step 07

    Decision recorded to audit

Fail closed. Decide at the edge. Record everything.

Developer overview