1. The fleet that had to coordinate without the network

Vernor Vinge’s A Fire Upon the Deep returns as the recalled work because its ships and minds operate across vast distances where messages take years and local decisions are final. The desktop is not crossing the galaxy, but it shares the constraint: there will be passes without ground contact, and during those passes the platform must remain coherent.

This entry applies the autonomy and FDIR literature to the desktop.

2. The desktop’s autonomy-relevant features

The desktop has several characteristics that shape its autonomy architecture:

  • Modular construction: cells, segments, and attachments can be added, removed, or fail independently.
  • Shared infrastructure: power, data, and thermal resources must be balanced across modules.
  • Long ground gaps: LEO orbits create regular communication windows, but also regular blackouts.
  • Mixed criticality: some payloads are revenue-generating and delay-tolerant; some functions are safety-critical.

These features push the desktop toward system-level autonomy with hierarchical FDIR.

3. Derived requirements

The autonomy literature gives the desktop six practical requirements:

  • Hierarchical FDIR: cells run local fault detection and recovery; the platform supervisor handles conflicts and resource reallocation. This matches the distributed-plus-coordinated model described in the literature.
  • Ground-outage operation: the platform must execute stored command sequences and contingency procedures for at least one full orbit without ground contact. The GOES-U Data Book cites seven days of autonomous operation as a design target for its bus; the desktop can start with a smaller but explicit autonomy horizon.
  • Graceful degradation: when a cell or attachment fails, the platform should shed non-critical loads, reconfigure networks and power paths, and preserve safe-mode functions.
  • Health telemetry aggregation: each cell reports a compact health summary. The platform fuses these into a single state that can be downlinked when contact resumes.
  • Model-based diagnosis where useful: for repeated, well-understood faults, model-based reasoning can reduce the need for ground-in-the-loop troubleshooting.
  • Safe mode as a deliberate state, not a panic: safe mode should preserve power, thermal balance, attitude, and communications readiness, and should be reversible when the fault is cleared.

4. Interaction with earlier arcs

The autonomy arc connects to several earlier decisions:

  • The fault recovery test arc (Entries 557–560) set the baseline for what the desktop must recover from.
  • The network reconfiguration test arc (Entries 477–480) established that the platform can reroute data paths.
  • The power distribution test arc (Entries 445–448) established that the platform can reroute energy.
  • The compute migration wondering arc (Entries 489–492) anticipated health-aware scheduling.

5. What this changes

  • The desktop inherits explicit autonomy and FDIR requirements.
  • Autonomy is not added on top; it is distributed through the existing cell-and-platform architecture.
  • The next entry will close the autonomy reading arc and record the decision.