ARGUS
ARGUSNOTE / 002REFERENCE MODELREVISION PENDING

A working reference model for the operational stack beneath autonomy.

The initiative treats observability, orchestration, intelligence, and supervision as operating primitives — properties of the system, not features of any single product.

CLASSREFERENCE ARTIFACT
STATUSIN STUDY
REVISION002.a
DOMAINAUTONOMY / INFRASTRUCTURE
NOTE / 002.0· WHY EXISTING MODELS FAIL

REFERENCE CONTEXT

CLASS
Infrastructure Theory
DOMAIN
Autonomy Systems
STATUS
In Study
REVISION
002.a
AUTHOR
ARGUS Initiative

Why most autonomy frameworks stop too early.

Most discussions surrounding autonomy focus on capabilities:

  • perception
  • planning
  • navigation
  • reasoning
  • execution

These are important, but they describe what an autonomous system can do — not how autonomous operations function at scale.

ARGUS studies the infrastructure layer beneath capability. The challenge is not whether a machine can act. The challenge is whether thousands of machines, agents, sites, and missions can be observed, coordinated, governed, and supervised simultaneously.

That operational layer remains largely undefined.

CENTRAL QUESTION
“The question is not whether an autonomous agent can act. The question is whether ten thousand autonomous agents can be understood, directed, audited, and controlled.”
— ARGUS WORKING NOTE 002
FIGURE 002.1

OPERATIONAL FLOW

Telemetry ascends. Authority descends. Operational state originates at the edge and moves upward through layers of interpretation, coordination, and supervision. Human accountability remains at the top of the system.

↑TELEMETRY
HUMAN AUTHORITYDecision Rights · Accountability · Strategic Control
SUPERVISIONOversight · Exception Handling · Governance
INTELLIGENCEInference · Prediction · Anomaly Detection
ORCHESTRATIONCoordination · Scheduling · Routing
OBSERVABILITYTelemetry · State Recovery · Indexing
OPERATIONAL SURFACEAgents · Machines · Infrastructure · Sensors
AUTHORITY↓
FIGURE 002.2

OPERATIONAL STACK

Five layers describe the operating substrate beneath fleet-scale autonomy — from physical agents at the edge to human authority at the top of the loop.

L4
SUPERVISIONHuman oversight, exception routing, accountability.
L3
INTELLIGENCEInference, anomaly surfaces, predictive signals.
L2
ORCHESTRATIONMission planning, scheduling, infrastructure routing.
L1
OBSERVABILITYTelemetry ingest, state reconstruction, indexing.
L0
OPERATIONAL SURFACEPhysical environments, agents, sensors, robotics, infrastructure, and edge operations.

DIAGRAM IS A WORKING ARTIFACT — SUBJECT TO REVISION AS THE STUDY DEVELOPS.

FIGURE 002.3

OPERATIONAL PRIMITIVES

The reference model is supported by a set of cross-cutting primitives. These are the dimensions along which any autonomous operation is observed, evaluated, and held accountable. They are not features; they are properties of the system.

P1
VISIBILITY
The recoverability of operational truth across time, agents, and environments.
  • What happened?
  • When?
  • Where?
  • Why?
P2
AWARENESS
The system's ability to transform operational state into actionable understanding.
  • What matters?
  • What changed?
  • What requires intervention?
P3
COORDINATION
The structured alignment of agents, missions, resources, and objectives across a shared operational picture.
P4
AUTHORITY
The allocation of decision rights throughout the operational stack.
  • Who may act?
  • Under what constraints?
  • With what accountability?
P5
LATENCY
The interval between event detection, interpretation, decision, and response.
P6
RESILIENCE
The system's capacity to maintain predictable behavior during disruption, degradation, and partial failure.

PRIMITIVES ARE OBSERVED ACROSS ALL LAYERS — L0 THROUGH L4.

REFERENCE MODEL · NOTE 002 · MAINTAINED BY THE ARGUS INITIATIVE

←RETURN TO ARCHIVE