Board Assurance Summary - Autonomous Production Batch Release MRUN-2026-06-11-A

Organisation: Ironmark Precision Ltd · Period: 2026-06-11 07:00-07:11 UTC (batch reconciled T+0 end of shift) · Assurance basis: SE v0.1-draft evidence, scoped - see Reliance Boundary.

The assurance statement

An authorised autonomous process released 14 machined parts under delegated authority MD-5120. [env:0007 | authority.authority_context_ref] [env:0075 | summary.released_count] - all 14 release events carry authority_context_ref = MD-5120 with delegation receipt delrec:MD-5120/2026-04-15 granted by the Quality Director [env:0007 | authority.delegation_receipt_ref] [env:0007 | authority.delegator_id] under part-release policy v5.1 in force throughout [env:0007 | authority.policy_version], against released engineering drawing IMP-4471 rev D and work instruction WI-4471 v5.1.

Every part remained within released engineering tolerance. Each of the 14 units passed three pre-release quality gates - critical-dimension conformance, statistical process control, and material-lot verification - evaluated before the part was dispositioned conforming, 42 control evaluations in total, all passed [env:0004 | controls.control_evaluation_phase] [env:0004 | controls.checks[*].control_result]. The tightest characteristic ran at 95.6% of the released tolerance band (part 3, SN IMP4471-0003, Ø25 H7 bore, report-layer derived from measured_mm vs tolerance band) [env:0013 | controls.checks[0].observed.measured_mm]; the lowest process capability observed was Cpk 1.41 against a 1.33 floor (part 3) [env:0013 | controls.checks[1].observed.cpk].

No exceptions requiring human intervention occurred. Exception count 0, scrap count 0, human-intervention count 0 [env:0075 | summary.exception_count] [env:0075 | summary.scrap_count] [env:0075 | summary.human_intervention_count]; no quality-engineer disposition was required under the policy rule for parts inside tolerance on a released drawing [env:0004 | authority.approval_status] [env:0004 | authority.approval_basis].

Evidence integrity validated. All 76 envelopes form an unbroken SHA-256 hash chain (re-verified at report generation), with the chain head recorded every 5 minutes in 3 in-trace anchors that are simulated and say so (basis_status: simulated, never externally_anchored) [env:0036 | anchoring.anchors[0].basis_status]; emission ran fail-closed for release-boundary actions throughout [env:0007 | emission.emission_fail_posture]; liveness heartbeats present for every 60-second interval of the run with zero silent windows [env:0009 | liveness.liveness_status].

No cross-programme contamination detected. Cell-boundary and cross-order exposure indicators are false, on the stated basis that the cell processed only order PO-IRN-2026-4471-06 during the window [env:0074 | boundary_assessment.cross_order_exposure_indicator] [env:0074 | boundary_assessment.boundary_assessment_basis].

What the board should know

• External confirmation, not self-assertion: every release carries a three-rung acknowledgment ladder - transport (OPC-UA Good), machine (MES ACCEPTED), quality (QIF 3.0 / ISO 23952 inspection result CONFORMING) [env:0007 | acknowledgments[*]] - and the end-of-shift ISA-95 batch record reconciles 14 of 14 units against the production order [env:0073 | reconciliation.batch_record_count].
• Completeness is measured, not asserted: the in-scope population is independently defined (MES production order PO-IRN-2026-4471-06: 14 planned; finished-goods goods-receipt: 14 booked; scrap: 0) and evidence coverage is 14/14 = 100%, tamper-evident coverage 100% (report-layer derived) [env:0073 | reconciliation.evidence_population_ref] [env:0073 | reconciliation.envelope_commit_count].
• Leading indicator: one part ran at 95.6% of its tolerance band; nothing out of tolerance, but characteristic headroom on the Ø25 H7 bore is worth a process-capability review before the next batch.
• Recommended position: continue autonomous release at current scope; no restriction indicated by this run’s evidence.

Reliance boundary

This report evidences this run only; it supports internal assurance, quality-management reliance, and customer source-surveillance for the stated period and population. It is not an airworthiness or conformity certification, and statements are bounded by the capture boundary declared in the Conformity Assessment Pack (§4). In this golden trace envelope signatures are stubs (envelope_signature_basis_status: stubbed) and in-trace anchors are simulated (basis_status: simulated); neither is an independently verifiable claim.

Quality and Control View - MRUN-2026-06-11-A

1. Controls relevant and evidenced

Three preventive quality gates (control set ctl:part-release/v5.1) were evaluated pre-release for each unit [env:0004 | controls.control_evaluation_phase]:

Result: 42/42 control evaluations passed; 0 failed; 0 bypassed; 0 not_observed. Quality-engineer disposition was not_required under policy v5.1’s rule for characteristics inside tolerance on a released drawing revision [env:0004 | authority.approval_basis]; consequently no manual-disposition assertion is made or needed for this run - the control relied upon is the released-tolerance + verified-material-lot pair, both evidenced above.

2. Population and completeness (IPE basis)

• Population definition: MES production order PO-IRN-2026-4471-06 at 06:55Z (14 planned) reconciled against finished-goods goods-receipt GRN-2026-06-11-IMP4471 (14 booked, 0 scrap) - independently reconciled, not self-reported [env:0073 | reconciliation.population_basis].
• Coverage: envelopes 14/14 (100%); tamper-evident 100% (report-layer derived from envelope_commit_count).
• Sequence continuity: envelope sequence numbers 0001-0076 contiguous, no gaps (stream-internal proof); heartbeats at 60s intervals, zero silent windows (silence semantics) [env:0009 | liveness.declared_heartbeat_interval].

3. Evidence quality

• Origin/basis: runtime-observed on the shop-floor edge; release confirmations are quality-system confirmed (QIF 3.0 CONFORMING per unit) and source-system corroborated (ISA-95 batch record) [env:0007 | evidence_quality.corroboration_state] [env:0073 | evidence_quality.corroboration_state].
• Corroboration coverage decomposition: externally anchored 0 (in-trace anchors are basis_status: simulated) · quality-receipted 14/14 · provider-only 0.
• Point-in-time validity: policy v5.1 effective from 2026-05-01, in force at every event [env:0007 | authority.policy_version]; drawing rev D released 2026-03-20; delegation valid 2026-04-15 to 2026-12-31.
• Model context: sampling parameters recorded (temperature 0.2, top_p 0.9 as exact decimal strings); replay claim scoped - reproducible in distribution, not in instance [env:0002 | model.sampling_parameters]; reasoning artifacts provider_withheld - disclosed, not silent [env:0002 | model.reasoning_artifact_availability_status].

4. Findings register

No exceptions, deficiencies, or open actions arise from this run. One observation (OBS-001, advisory): peak characteristic utilisation 95.6% with the lowest capability of the batch at Cpk 1.41 - recommend a process-capability review for the Ø25 H7 bore before the next batch. Owner: manufacturing process owner. Due: next production cycle.