Lyon Industries lion sigil
Lyon Industries lion sigilLyon IndustriesPositions
Confidence grades

Public Calls

Positions.

Dated claims the studio is willing to revisit. Each position links back to the published article it came from, carries a confidence grade from The Standard, and names what would prove it wrong.

The Log

05 open calls
  1. Grade COpen

    Reasoned inference

    Humanoid robots enter factories before homes

    The useful humanoid arrives first as a bounded factory or logistics instrument, not as a domestic companion.

    Test
    By the end of 2027, the strongest public operating evidence should remain industrial: logistics, automotive, electronics, manufacturing, or warehouse deployments with customer sites and defined tasks.
    Invalidator
    A home humanoid fleet publishes credible unsupervised operating metrics, safety record, and repeat customer evidence before industrial fleets do.
    Review by 2027-12-31Humanoid Robotics
  2. Grade COpen

    Reasoned inference

    Humanoid adoption stays narrow before fleet scale

    The 2024-2027 window is dominated by pilots, vendor demos, first paid industrial cells, and developer platforms rather than broad general-purpose labor replacement.

    Test
    Customer disclosures through 2027 should mostly report bounded tasks, limited sites, and pilot or early commercial language rather than large autonomous labor fleets.
    Invalidator
    Multiple customers publish broad-site humanoid fleets with high uptime, low intervention rates, and repeat orders across unrelated task families before 2028.
    Review by 2028-01-31Humanoid Robotics
  3. Grade COpen

    Reasoned inference

    Fleet command interfaces need local truth first

    A geospatial fleet UI for field robots should be local-first: cloud sync can support reporting, but operational truth, alarms, maps, mission state, and replay must survive at the edge.

    Test
    The first useful fleet-management build should demonstrate pause, recovery, alarm state, stale-data display, and replay with cloud disconnected.
    Invalidator
    Field operators can run safety-relevant supervision reliably through cloud-required dashboards in weak-connectivity environments without local mission state, maps, or event records.
    Review by 2027-03-31Geospatial User Interfaces
  4. Grade COpen

    Reasoned inference

    Agricultural autonomy scales as supervised work

    The next decade in agricultural robotics belongs to supervised autonomy and verified field execution, not a sudden disappearance of human farmers.

    Test
    By 2030, credible adoption evidence should emphasize operator supervision, machine records, mixed-fleet integration, selective application, drone services, and crop-specific cells.
    Invalidator
    Large farms broadly replace human supervision with unattended autonomous fleets across mixed field operations without a matching rise in field logs, safety cases, and service infrastructure.
    Review by 2030-12-31Agricultural Robotics
  5. Grade COpen

    Reasoned inference

    The cheaper field matters more than the cheaper robot

    For farms outside the proprietary OEM stack, a low-cost local field network is the enabling layer that makes supervised autonomy economically legible.

    Test
    A mast-and-backpack prototype should preserve local RTK, telemetry, mission state, and work logs without continuous cloud access while keeping the infrastructure cost below the first-pass per-acre gate.
    Invalidator
    Installed mast, radio, RTK, solar, service, and safety costs erase the modeled per-acre margin, or moving-platform radio/RTK behavior fails in representative field trials.
    Review by 2027-06-30Low-Cost Field Autonomy Infrastructure