Marginally ahead of pace, seven quarters into a three-year commitment.
Scorecard — the whole state in twenty seconds.
Level 0 · metric treeEvery number below reappears somewhere deeper in this report with its evidence attached. The headline is Balance — the share of in-scope line items sitting inside their balanced band: above the safety-stock floor, below the max/MOQ-adjusted ceiling.
The understock uptick since Q4 2027 (7.0% → 7.8%) is the mill-circuit criticality restatements raising floors, not stock disappearing — logged in §02, priced in §03.
Measure summary Original · current · target · status
| Measure | Original — Q3 2026 | Current — Q2 2028 | Target — Q2 2029 | Status |
|---|---|---|---|---|
| In-scope stock value | $52.4M | $47.1M | $44.0M | On track |
| Cumulative release vs pace | $0 | $5.3M | $8.4M | On track |
| Imbalanced — items outside the band | 37.0% | 23.6% | <15% | On track |
| Overstock — above ceiling | 26.1% | 15.8% | <10% | On track |
| Aged tranche · >5yr stock-on-hand | $4.9M | $3.1M | <$1.5M | At risk |
| Overstock with open orders | 4.0% | 2.1% | <0.5% | At risk |
| Understock — below floor | 10.9% | 7.8% | <5% | At risk |
| Vital-tier understock | 2.4% | 0.9% | <1% | On track |
| Understock with no open orders | 5.1% | 3.4% | <1.5% | Behind |
| Reviewed — program penetration | 0% | 64% | 100% | On track |
| Fix approved — of imbalanced items | 0% | 41% | ≥50% | Behind |
Bar = progress from original to target; the tick is where the bar should be at quarter 7 of 12. At-risk rows are within 10pts of pace; the understock slippage traces to the mill-circuit restatements (§02) — floors rose faster than orders were raised; the fix is this quarter's Vital-tier increases (§03).
Ahead of the glidepath since Q3 2027. The aged tranche is the drag — it clears by disposal, not policy.
Floors rose with the mill-circuit restatements and orders haven't caught up — 820 items have nothing inbound. This quarter's Vital-tier increases are the fix.
The program — past, present and future against a fixed line.
3-year arcAt program start, in-scope MRO stock stood at $52.4M. ABC committed to $44.0M by Q2 2029 — an $8.4M working-capital release, net of the protective increases the program requires. The baseline and target move only by logged restatement, never by drift.
Restatement log Governance
Performance and restatement are always reported as separate lines — reclassification that changes the optimal is logged openly, so truth-seeking never games the target.
| Quarter | Restatement | Items | Reason | Effect on target |
|---|---|---|---|---|
| Q2 2028 | Criticality corrections — mill circuit | 38 | Post-failure VEDN review raised 38 floors; optimal and target restated | ▲ +$0.21M |
| Q4 2027 | Lead-time class corrections | 12 | Vendor 100482 re-banded after sustained 3× lead-time inflation | ▲ +$0.06M |
| Q2 2027 | Duplicate material merge | 21 | Cleanse-driven consolidation reduced optimal holdings | ▼ −$0.09M |
Stated together, this quarter: balance improved +1.3pts (performance); target moved +$0.21M via 38 criticality restatements (restatement). Never blended.
The quarter — what Q2 2028 contributes.
Working detailThis quarter's run prices 10,240 candidate changes and triages them into a workload that fits ABC's ~240 review-hours. Its recommendation set is worth −$3.9M net stock value and −$2.4M/yr in expected cost — two ledgers, reported separately throughout.
The closeness of those two lines is what makes this quarter's forecast credible. When they diverge, the divergence is reported here first.
The quarter in review
The sore spot — the mill train. After ML-02's trunnion bearing failure in February (11 days down, the spare expedited from Perth at triple freight), the reliability team pushed criticality reviews on 340 mill-circuit materials. This quarter's simulation is the first to price those revised criticalities — it's why the Vital-tier increases ($480k) are the largest they've been in the program. Approving them is the cheapest insurance ABC buys this year.
The opportunity — the gold room and tails circuit. Elution heaters, carbon screens and tails pumps were stocked in 2019 for a second CIL train that never cleared feasibility. Demand history now shows them cycling at half the assumed rate; they carry ~$1.1M of the reduction opportunity on their own, and none of it touches a vital spare.
The quiet win. Emergency freight on actioned items fell 38% year-on-year — the Q3 2027 approvals working as simulated. The remaining emergency spend concentrates on the unreviewed backlog, which is the strongest argument for clearing it: the register's oldest untouched items are now its most expensive habit.
The register — the working level, per lane.
Review workflowEvery row carries a review state that persists and rolls forward — the backlog is the program's memory. High-stakes rows link to their dossier (§05), the evidence file behind each recommendation.
High-stakes lane 380 items · individual review · ~210 hrs
| Material | Description | Unit price | VEDN | Movement | Current min/max | Rec. min/max | Δ Stock value | Δ Annual cost | Driver | Conf. | State |
|---|
Reading a row: Δ Stock value is balance-sheet (an increase costs capital); Δ Annual cost is P&L (the recurring cost it avoids) — never netted against each other. Confidence is backtest agreement; Med rows are where the model wants a human.
Bulk lane 7,740 items · approve-by-batch with exception flagging
Quarantine 1,180 items · shown, reasoned, not recommended
The credibility lane — the model naming what it won't judge is what makes the other lanes trustable.
The dossier — the evidence file, one per high-stakes item.
Template · shown onceEach of the 380 high-stakes items carries this four-chart file. The reviewer sees the demand, the root cause, the simulated economics, and the sign-off table — no figure is model-generated without its evidence shown.
Gearbox input shaft, CR-East
Hold one more unit. Lead time has tripled on a vital spare protecting a chronic asset; the extra unit's holding cost is trivial against the downtime it prevents.
Provenance — demand from MB51 (41mo); lead times from EKBE; asset linkage from IW39; simulation at the Vital service target (98%) from the agreed consequence baseline; forecast backtested.
Appendix — terms, definitions, boundaries.
Baseline $52.4M (Q3 2026) · target $44.0M ±0.5 by Q2 2029 · service levels per VEDN tier (V 98% · E 95% · D 92% · N 88%) · carrying rate 18%/yr. Set jointly at program start; fixed except by logged restatement (§02).
Floor = safety stock at the tier service level. Ceiling = max, or MOQ-adjusted receipt peak, whichever is higher. Items whose policy changed within the run-down window are in-transition: excluded from the headline, shown as the converging count.
MB51 (issues, 41mo) · EKBE (receipts / lead times) · MARC/MBEW (policy, valuation) · IW39 (work orders / asset linkage) · ME2M (open POs). Read-only; refreshed 04 Jul 2028.
Probabilistic forecasting (backtested), Monte-Carlo policy simulation (10,000 years per item-policy), triage funnel per §03. Sparse or contradictory items are quarantined, not guessed.
The two ledgers are never blended. Item-level stockout avoidance is never claimed — only cohort divergence. Structural / MOQ-driven stock is attributed to procurement terms, not policy failure. No figure is model-generated without shown evidence.
Scorecard (dense) → program (spacious) → quarter (working) → register (tables) → dossier (evidence). One temporal grammar throughout: solid past, filled-dot present, dashed forecast, amber-ring target.