Milestone Gating
Commitments are structured so that each phase must earn evidence before the next phase proceeds.
Gate before you go
Milestone gating is a decision discipline: a program does not proceed to the next phase until the current phase has produced defined evidence. Each gate has entry criteria, a body of evidence, and an explicit go/no-go. This protects a defense counterpart from a program that spends momentum instead of proof.
What a gate requires to pass
- Defined, falsifiable entry and exit criteria written before the work, not after.
- Evidence matched to the claim — hardware for hardware, reproducible code for simulation.
- An explicit decision recorded, including no-go and rework outcomes.
- No borrowing of confidence from a later phase to pass an earlier one.
Why milestone-gating fits fusion
Fusion programs fail slowly and expensively when phases blur together and unmet assumptions carry forward. Gating forces the open questions — net tritium breeding, the plug-density requirement, FOAK integrated operation — to be confronted at defined points rather than deferred indefinitely. The FOAK-NOAK-BOAK build sequence is itself a set of gates: each unit must operate before the next is trusted.
Milestone gating is also how Kronos keeps its public claims aligned with its evidence: a claim is upgraded only when the gate that justifies it is passed.