1. The station that ran on unanswered questions

Frederik Pohl’s Gateway is about an asteroid base full of alien ships that humans barely understand. The crews who fly those ships know the odds are bad, but they go anyway. What makes the novel memorable is not the ships; it is the operation around them — the psychologists, the gamblers, the bookies, the administrators, and the computers that all try to manage a venture no one fully comprehends. The desktop is not a death trap, but it is similarly a machine that must be operated before it is fully understood.

Entry 264 closed the ADCS arc. This entry opens the operations arc by asking how the desktop will be run once it is in orbit.

2. What operations means

Operations is everything that happens after launch to keep the desktop useful. It includes:

  • Ground segment: the antennas, networks, and computers that communicate with the spacecraft.
  • Mission control: the people and software that plan, command, and monitor the desktop.
  • Flight software: the onboard programs that execute commands, run autonomy, and report telemetry.
  • Procedures: the written and automated sequences for normal and emergency situations.
  • Data flow: the path from sensor to customer, and from customer request to command.

Without operations, the desktop is just hardware in orbit. Operations is what turns the hardware into a service.

3. The operations questions

The operations arc should answer:

  • Who talks to the desktop, and from where?
  • How often can the desktop be contacted, and for how long each time?
  • What decisions does the desktop make on its own, and what decisions need a human?
  • How are commands planned, verified, and sent?
  • How is telemetry stored, downlinked, and acted upon?
  • What happens when contact is lost?
  • How are customer requests translated into spacecraft commands?

These questions are not specific to the desktop; every spacecraft must answer them. But the desktop’s answers will shape its cost, its staffing, and its customer experience.

4. Ground segment options

There are several ways to operate a LEO spacecraft:

  • Own ground stations: build or lease antennas in a few locations. Gives control but requires capital and staff.
  • Ground station networks: use commercial networks like KSAT, SSC, or Atlas. Pay per pass or per minute.
  • Relay networks: use systems like Starlink, Iridium, or Inmarsat for near-continuous connectivity. Higher data cost but simpler operations.
  • Hybrid: use owned or contracted S-band stations for command and telemetry, and a relay network for high-rate payload data.

For the desktop, a hybrid is likely. S-band or UHF ground contacts for command and housekeeping, and a commercial relay or ground network for payload data when the bandwidth demand justifies it.

5. Autonomy level

The desktop must decide how much to do without human approval. The extremes are:

  • High autonomy: the desktop schedules its own observations, handles safe mode, and only reports summaries to the ground.
  • Low autonomy: every observation, mode transition, and manoeuvre is commanded from the ground.

Most small satellites operate somewhere in between. Routine housekeeping is autonomous. Payload operations are pre-planned and uploaded as command sequences. Anomalies trigger safe mode and a call for help.

For the desktop, the target is moderate autonomy. It should be able to survive for days without contact, maintain power and thermal safety, and execute a stored schedule. Complex customer tasks or propulsive manoeuvres are planned on the ground and uploaded.

6. Command and telemetry

Commands are the instructions sent to the desktop. Telemetry is the data sent back. The command set should be small and explicit:

  • Mode commands: enter safe mode, enter nadir pointing, enter Sun pointing.
  • Payload commands: start imaging, stop imaging, configure sensor.
  • Propulsion commands: fire thruster set A for duration X.
  • Configuration commands: set wheel speed limit, update controller gains.
  • File management: delete file, downlink file, list files.

Telemetry should include:

  • Housekeeping: voltages, currents, temperatures, wheel speeds, attitude estimate.
  • Payload status: imager state, image count, storage use.
  • Event logs: mode transitions, fault flags, command history.
  • Payload data: images and derived products, either in real time or stored for later downlink.

Commands and telemetry are usually packaged in a protocol like CCSDS and transmitted over a radio link. The desktop’s compute attachment handles parsing, execution, and storage.

7. The operations timeline

A typical day in the desktop’s life might look like:

  1. Overnight autonomous operations: maintain Sun pointing, charge batteries, run scheduled imaging or data processing.
  2. Morning ground pass: downlink housekeeping and stored data, upload the next day’s schedule.
  3. Daytime operations: execute the uploaded schedule, take images, process data, respond to customer requests.
  4. Evening ground pass: downlink priority data, confirm schedule execution.
  5. Repeat.

This rhythm assumes two to four ground contacts per day. With a relay network, the rhythm becomes more continuous but also more expensive.

8. Anomaly response

When something goes wrong, operations must know what to do. The levels of response are:

  • Autonomous recovery: the onboard software detects a fault and fixes it, for example by switching to safe mode.
  • Ground-instructed recovery: the operator receives a fault flag and sends a corrective command.
  • Procedure-based recovery: a known anomaly has a documented response sequence.
  • Investigation: an unknown anomaly requires analysis before any action is taken.

The desktop should have a small library of procedures for common faults: safe mode entry and exit, wheel saturation, star tracker blinded by Sun, communications loss, power bus undervoltage, thermal limit exceeded.

9. Operations cost

Operations is often underestimated. It includes:

  • Ground station fees or capital costs.
  • Mission control staffing, either in-house or contracted.
  • Flight software maintenance and updates.
  • Telemetry storage and processing on the ground.
  • Customer interfaces and order management.

For a first desktop, operations might be 20–40 percent of the total mission cost over five years. As the fleet grows, the marginal cost per spacecraft falls because the same ground segment and team can support multiple units.

What this changes

  • Operations is the next attachment to define, following ADCS.
  • The desktop will likely use a hybrid ground segment with S-band or UHF contacts plus commercial relay or network services for data.
  • Target autonomy is moderate: survive days without contact, execute stored schedules, escalate complex decisions to the ground.
  • Commands and telemetry need clear protocols and a small, well-defined command set.
  • Anomaly response ranges from autonomous recovery to ground-instructed procedures.
  • The next entry will size the ground segment and communications cadence.