Given a launch-cost collapse but persistent uncertainty about which orbital industries will scale, what concrete experimental or contracting designs (e.g., options on future capacity, milestone-based offtake, pre-competitive shared testbeds) would most efficiently discover where Wright’s-law learning curves are steepest—across orbital manufacturing, compute, and servicing—without stranding large amounts of overbuilt, underused hardware?

starship-orbital-economy | Updated at

Answer

Use small, time-bounded, multi-tenant experiments with staged commitments and strong backpressure from utilization data. Mix options, milestone offtake, and shared testbeds so learning scales while capex stays modular.

  1. Contract structures

  • Capacity options

    • Customers pay small premia for rights to future orbital capacity (rack-space, robot-hours, compute) on specific shells/platform types.
    • Exercise windows tied to tech milestones or demand triggers.
    • Lets operators size hardware in tranches as options are exercised; reveals where willingness to pay clusters.

  • Milestone-based offtake

    • Buyers commit to buy X campaigns or compute-hours if provider hits objective metrics (yield, latency, uptime, safety certs).
    • Milestones stair-step: demo → pilot scale → semi-commercial.
    • Shifts risk to providers but gives clear price/volume signals by industry.

  • Portfolio “sandbox” agreements

    • Agencies or large firms fund a basket of small orbital trials (manufacturing, compute, servicing) with preset down-select rules.
    • Fixed total budget, rotating calls; only winners get scale-up offtake.
    • Encourages parallel probes without locking into a single thesis.
  1. Experimental platform designs

  • Modular, campaign-based testbeds

    • Standard racks or pods with fixed campaign durations (e.g., 3–12 months) sold as slots.
    • Rotating tenant mix across manufacturing, compute, and servicing demos.
    • High reuse of bus, power, GNC; hardware growth tied to slot utilization.

  • Shared servicing sandboxes

    • Small clusters of standardized targets, fixtures, and dummy assets in one shell.
    • Multiple tug/robot providers run trials against common tasks and metrics.
    • Learning curves measured per docking, per inspection, per kg maneuvered.

  • Pop-up micro-factories with return capsules

    • Short-lived, single-purpose units for candidate microgravity products.
    • Standard bus + process-specific payload; strict deorbit.
    • Lets many product lines be tested without building a large, permanent factory.
  1. Cross-industry learning and Wright’s law

  • Common metrics and telemetry

    • Require simple, shared KPIs: cost per robot-hour, per rack-kWh, per processed kg, per maneuver.
    • Publish anonymized curves where possible to steer demand toward steepest declines.

  • Multi-tenant industrial shells

    • Cluster trials in a few orbits with shared depots and tugs.
    • Volume across sectors (manufacturing, compute, servicing) drives common hardware and ops learning.
  1. Overbuild avoidance tactics

  • Hard utilization gates

    • New platform mass only launched when prior tranche hits utilization/price targets.
    • Contracts specify automatic slowdown if slot fill or price realization lags.

  • “Design-to-sunset” assets

    • Default mission profiles: cheap, short-lived hardware that fully deorbits.
    • Persistent platforms reserved for use cases that pass strong demand and learning thresholds.

  • Rights-constrained experiments

    • Treat orbital rights (slots, debris budget) as explicit cost in pilot design.
    • Forces early business models to internalize congestion risk before scaling.
  1. Who runs this
  • Public–private testbed operators
    • Agencies anchor basic platforms and data standards; private firms sell capacity.
  • Consortia for pre-competitive R&D
    • Shared manufacturing and servicing trials where IP is process-level, not platform-level.

Overall: start with small, standardized, campaign-based platforms and option-like commitments; scale only where repeated campaigns show fast cost decline and strong willingness to pay, letting weak lines die without leaving large, idle stations or factories.