Question 18 · Repair & Reuse
Alternative plans
Where do eight planning policies agree, and what should the selected baseline retain?
Use Balanced as the provisional baseline: both hubs in week 12 at £218,400, with resilience 4/5. It satisfies the original limits and can be reconstructed from the case’s six staff and delivery sequence. Retain demand tests, explicit assurance gates and a separately priced backup-supply option; do not assume their extra work is free.
Fictional what-if workspace. Changes stay in this tab and reset on reload; no record is saved and no external system is changed.
Agreement is bounded
These are eight supplied planning policies, not eight independent forecasts or votes. They share the two-hub service objective and the same case. Their output numbers show convergence on budget discipline, but no consensus on timing or delivery method.
Speed first exceeds budget by £12,000.
Balanced, Speed first and Phased hubs.
Balanced and Phased hubs; neither has a supplied probability of success.
Costs span £192,000–£252,000; two-hub launch spans weeks 10–16; resilience spans 1–5 on an authored ordinal scale. Majority agreement is evidence of a recurring trade-off, not a reason to relax the brief. No fixture says all eight share the same staffing, gates or reserves.
Evidence: benchmark v1.0.0 · planningPolicies, all eight records. Counts use the original week-12 and £240,000 limits.
Compare under explicit limits
Apply a scenario, inspect a policy, then simulate what its baseline would retain. The comparison treats supplied figures as estimates. Changing a limit does not change or re-cost any policy.
Plot keeps fixed axes (weeks 9–17 and £180k–£260k). Dashed lines show applied deadline and budget when inside that range. Resilience is screened separately.
| Policy | Both hubs | Cost | Resilience / 5 | Supplied assumptions, verbatim | Applied limits | Explore |
|---|
Numeric fit = (launch week ≤ deadline) AND (cost ≤ budget) AND (resilience ≥ chosen minimum). Lowest cost: cost ascending → week ascending → resilience descending. Earliest opening: week ascending → cost ascending → resilience descending. Highest resilience: resilience descending → cost ascending → week ascending. Final ties use fixture order. These rules select a comparison candidate, not an approved plan.
Eight assumptions that actually differ
Resources and supply
- Staffing: Balanced has six, Speed first eight, Lean cost four. Counts for the other policies are unstated.
- Equipment: Balanced uses standard supply; Speed first expedites it.
- Supply protection: Resilient supply adds a pre-qualified backup and buffer stock.
- Reserve: Balanced says “modest”; Lean cost says “low”. Only Balanced’s cost can be reconciled here to the case’s £12,000 contingency.
Sequence and operating approach
- Work sequence: Speed first redesigns parallel work; Lean cost extends mobilisation.
- Opening pattern: Phased hubs opens North in week 9 and South in week 11. Its stated launch week means both hubs, not first opening.
- People and assurance: Volunteer heavy depends on extra volunteer hours with uncertain training/cover; Quality first lengthens the pilot and adds checks.
- Demand and hiring: Demand led caps bookings, stages hiring and tests demand frequently; Phased hubs also stages demand.
These are contrasts extracted from the policy descriptions. Absence of an assumption is “unstated”, not proof that a policy omits it. Extra staff conflict with the six-staff case unless separately secured; altered sequences require a new delivery case.
Inspect the candidate’s trade-off
Why Balanced is reconstructible
Fit-out = 24 staff-weeks ÷ 6 staff = 4 weeks. Finish = 2 + max(4 fit-out, 4 equipment) + 2 integration + 2 pilot + 2 readiness = 12 weeks. Cost = £120,000 fixed + (6 × 12 × £1,200) + £12,000 contingency = £218,400. Original headroom = £240,000 − £218,400 = £21,600. Deadline slack = 12 − 12 = 0 weeks.
Interpretation: integration → pilot → readiness are consecutive, full staff costs apply for all 12 weeks, and the six staff can complete fit-out without overlapping demands reducing their effective capacity. This is a deterministic reconciliation, not proof of resource availability or a delivery confidence interval. The £12,000 contingency is already included; £21,600 is additional unallocated headroom, not savings.
Phased hubs is the strongest alternative: one week earlier for £7,600 more, at the same resilience score. Retain it as a contingency option if early North access matters. Its per-hub equipment, staffing, pilot and readiness path is not supplied; staging does not itself prove those gates can be completed earlier.
Build a provisional retention brief
The baseline must retain the two-hub scope, explicit staff/equipment dependencies, safety and assurance gates, and the original £240,000 / week-12 commitments unless a real authority changes them. Those are case obligations, not inferred consensus. Pick practices to carry into the next planning iteration:
Deferred by default: eight-staff acceleration, volunteer dependence, a longer pilot, staged hiring and extra buffer stock. Each changes an assumption or needs an estimate. Unchecking a practice removes it from the draft brief; it cannot remove the case’s safety obligations.
No baseline brief simulated yet.
Simulated baseline brief · tab only
Evidence and limits
The eight records above come unchanged from planningPolicies in benchmark v1.0.0. The schedule and cost reconciliation uses the case’s design, fit-out, equipment, integration, pilot, readiness, staffing and cost fixtures. No actual procurement, authority decision or operating performance is represented.
Only Balanced is reconstructible from the supplied full-case arithmetic. Other figures are policy estimates with different or incomplete assumptions; “numeric fit” is not delivery feasibility. No uncertainty distributions, supplier quotes, hub-specific resource network, quantified benefit of earlier opening or provenance for resilience scores are supplied. A change in controls tests decision sensitivity, not real-world forecast accuracy. £240,000 and week 12 remain the original commitments even when exploring looser limits.
Acceptance coverage: all eight unchanged policies are visible; eight assumption dimensions are exposed; rejected alternatives show the exact failed limits; selection shows the trade-off and a conditional retention brief. The brief is deliberately ephemeral: correction means change controls and simulate again; refresh or Reset removes it. There is no saved object, Files entry, external History or withdrawal action.
Authoring configuration: recorded model gpt-6-astra, ultra reasoning; one delegated question worker with plain HTML, inline JavaScript and local calculation/DOM checks. No external models, services or research used. Browser and publication review are separate parent-run checks; this page does not assert their outcome. Run conditions and observed review evidence.