Given Starship-scale launch-cost collapse and early adoption of shared orbital interfaces, what specific quantitative thresholds in traffic density and platform reuse (e.g., launches/year, standardized-rack turns/year) are needed for Wright’s-law effects to drive simultaneous cost crossover in at least two distinct industries—such as orbital compute and microgravity manufacturing—rather than each sector having to climb its own separate learning curve?

starship-orbital-economy | Updated at

Answer

Plausible but uncertain thresholds (order-of-magnitude, not precise targets):

  1. Shared launch & traffic density
  • Starship-class launches to LEO: ~150–300 launches/year on a single vehicle family, with ≥50–70% of flights serving mixed payloads (compute+manufacturing+others).
  • Cumulative flights for strong Wright’s-law effects in launch ops: ~1,000–2,000 total flights of the family.
  • Result: launch integration, ops, and basic rideshare interfaces ride a single learning curve used by both industries.
  1. Standardized rack / module scale Assume a common “container” (rack or small free-flyer bus) used for both orbital compute and microgravity modules.
  • Active standardized racks in orbit (all customers):
    • Early crossover band: ~3,000–10,000 racks total, shared across industries.
  • Rack turns (install/remove or major reconfig):
    • ~3–6 turns/rack/year for compute,
    • ~1–3 turns/rack/year for manufacturing lines and experiments.
  • Cumulative standardized-rack deployments (including swaps and refurb):
    • Lower bound for strong Wright’s-law in rack hardware and handling: ~50,000–100,000 cumulative rack-equivalents.

At these levels, design, production, integration, and ops of racks, power/data interfaces, and robotic handling see learning that benefits both industries simultaneously.

  1. Platform and servicing reuse Assume a small family of uncrewed platform buses and a common servicing stack.
  • Multi-tenant platforms (shared buses) in a given shell: ~50–150 platforms using the same power/data/mech interfaces.
  • Average platform life with significant payload churn: ≥8–10 years, with ≥3–5 full payload refresh cycles.
  • Cumulative standardized docking/berthing/robotic service events: ~10,000–30,000 events using common fixtures.

This density lets servicing robots, tugs, and ops software amortize their learning across both compute and manufacturing payloads.

  1. Simultaneous cost-crossover conditions (very approximate) For both orbital compute and at least one microgravity-manufacturing product to hit cost crossover using shared learning:
  • Launch cost: sustained ≤$100–300/kg to the relevant LEO shell (Starship-scale) with high reuse.
  • Effective “all-in” rack deployment cost (hardware + prep + launch + install):
    • Falls by ~60–80% from first-generation values after ~50k–100k rack-equivalents.
  • Servicing / robot-hour cost in orbit (for standardized tasks):
    • Falls by ~50–70% after ~10k–30k standardized service events.

Under these conditions:

  • Orbital compute: can become competitive first on high-value secure or rad-hardened workloads once effective $/kW-month (including capex amortized over many similar racks) approaches 1.5–3× premium over top-tier terrestrial secure data centers.
  • Microgravity manufacturing: at least one product with very high value/kg (e.g., specialty fiber or pharma intermediate) can cross over once shared rack and servicing learning remove most bespoke integration cost and leave product-specific process as the main premium.
  1. Why simultaneous crossover depends on shared standards These thresholds only drive joint crossover if:
  • ≥70–80% of launches, racks, and servicing events use the same families of:
    • mechanical mounts,
    • power/data connectors and protocols,
    • grapple/docking fixtures,
    • ops software APIs.
  • Both industries accept some suboptimal tailoring to stay inside these standards.

Otherwise, each sector’s effective cumulative volume (for Wright’s law) fragments and they climb separate, slower curves even at similar traffic densities.

  1. Practical near-term target band Given uncertainties, a “pragmatic” band where cross-industry learning has a real chance to matter is:
  • ~100–200 Starship-class launches/year with strong rideshare.
  • ~5,000–20,000 standardized racks deployed cumulatively.
  • ~5,000–15,000 standardized servicing/robotic handling events.

Below this band, simultaneous crossovers for two distinct industries are unlikely; above it, the binding constraint becomes demand and product-market fit rather than cost from launch/platform/servicing per se.