04 System Architecture

Icarus Orbital Systems — System Architecture

The technical architecture · three components · operational model. Source for all pitch content.

Three-component fleet architecture

graph TB
    subgraph SPACE["LEO Operations · ELEO 200-2000km"]
        PTOT[Primary Tele-Operated Tug<br/>Command + Control Hub]
        DF1[Dragonfly DF-1 Drone Flotilla<br/>Active capture · 6-DOF]
        OPS[Orbital Processing Stations<br/>Phase 2 · sort + refine]

        PTOT -->|directs| DF1
        DF1 -->|captures| DEBRIS[(Orbital debris<br/>1U CubeSat → fragments)]
        DF1 -->|transfers| PTOT
        PTOT -->|stages| OPS
        OPS -->|recovered material| SUPPLY[In-orbit supply chain]
        DEBRIS -.->|controlled deorbit alt| EARTH[Earth atmosphere]
    end

    subgraph GROUND["Ground Operations"]
        OPS_CTR[Tele-op + AI assist console]
        OPS_CTR -.->|commands| PTOT
    end

    style PTOT fill:#E0B25E,stroke:#B8893A,color:#0B1120
    style DF1 fill:#B8893A,stroke:#8B5E3C,color:#fff
    style OPS fill:#4FD8FF,stroke:#2B7AB8,color:#0B1120
    style SUPPLY fill:#3D7A4E,stroke:#1F4D4A,color:#fff

Component A: Primary Tele-Operated Tug (PTOT)

Role: Command-and-control hub.

Attribute Spec
Operation mode Human-supervised + AI-assisted
Crew None (uncrewed; ground-controlled)
Function Coordinates Dragonfly fleet · stages captured material · interfaces with OPS
Direct debris contact No — directs, doesn’t capture
Loss profile Centralized · not in contact zone

Component B: Dragonfly DF-1 drone flotilla

Role: Active debris interaction.

Attribute Spec
Configuration Modular · distributed fleet · 6-DOF maneuverability
Propulsion Hall-effect (xenon) primary + hydrogen peroxide RCS thrusters
Capture systems Articulated end-effector (stable targets) + deployable nets (poly-mylar + carbon-strand for tumbling fragments)
Target range 1U CubeSat → large fragments
Resilience Distributed — loss of one unit = degraded ops, not failure

Component C: Orbital Processing Stations (OPS) [Phase 2]

Role: Sort · refine · stage recovered material for in-orbit reuse or controlled deorbit.

Attribute Spec
Location ELEO (Equatorial Low Earth Orbit) · 200-2000km altitude · ±5° latitude
Why ELEO Geometry minimizes inclination changes between debris-dense zones
Output Recovered metals · composites · fed back to in-orbit supply chain
Phase 2 (after demonstration mission validates capture economics)

Operational model — Orbital Dredging

flowchart LR
    A[1. Identify debris-dense corridor] --> B[2. Position PTOT]
    B --> C[3. Deploy Dragonfly flotilla]
    C --> D[4. Capture · stabilize · transfer]
    D --> E{5. Process or deorbit?}
    E -->|process| F[OPS · refine · stage]
    E -->|deorbit| G[Controlled re-entry]
    F --> H[Material recovery revenue]
    D --> I[Removal contract revenue]
    style I fill:#3D7A4E,stroke:#1F4D4A,color:#fff
    style H fill:#3D7A4E,stroke:#1F4D4A,color:#fff

Each step is independently revenue-generating. Phase-1 economics (debris removal) close before OPS comes online.

Why distributed beats monolithic

Architecture choice Pro Con
Distributed Dragonfly fleet (chosen) Loss of 1 unit = degraded ops · scalable · parallel work More units to manufacture · coordination overhead
Single large tug (alternative) Simpler ops · one mission · centralized control Single point of failure · serial work · risk concentration

Distributed wins for resilience + parallelism. Coordination overhead solved via PTOT + AI-assist.

Why tele-op + AI assist beats fully autonomous

Architecture choice Pro Con
Tele-op + AI assist (chosen) Insurance approval · regulatory posture · human-in-loop trust Slower than fully autonomous · ground crew cost
Fully autonomous Faster cycles · no comms latency Insurance · regulatory · fail-mode liability open question

Tele-op + AI gets us to operations sooner. Fully autonomous is a Phase-3 transition, not a Day-1 decision.

Cross-references