A Requirement Gap That Presents as a Schedule Problem
You inherit a 240-unit sensor retrofit across 12 sites, planned for 20 weeks. Requirement R-14 requires 100% functional test of every unit before shipment; R-31 caps site labour at 2.0 technician-hours per unit. Pilot data from 18 units: functional test averages 2.6 h per unit at the depot and identifies 7% of units as defective, each needing about 5 h of rework. Whether the test belongs in site scope was never analysed; the pilot simply ran it at the depot. Three sites lose access after week 18. The client's schedule penalty is $4,800 per day beyond week 20. The requirements baseline was signed seven months ago with no test-location decision recorded. A delivery lead proposes to test everything at the depot and absorb it in the contingency.
1. Is this a schedule problem or a requirements problem? Show the analysis that decides it.
Read the worked answer
It presents as schedule but it is a requirements problem surfacing as schedule. R-14 fixes what (100% test before shipment). R-31 fixes a resource envelope at the site. Neither assigns a location, and the baseline has a gap rather than two contradictory statements - which matters, because a gap is resolved by analysis and a recorded decision, while a contradiction is resolved by a change to one of the requirements.
Do the capacity math before choosing an option. Test work: 240 units × 2.6 h = 624 h, plus rework of 7% × 240 ≈ 17 units × 5 h ≈ 85 h, so roughly 709 h of test-and-rework content. R-31's total site envelope is 240 × 2.0 = 480 h for everything the site does. The test cannot fit inside R-31's envelope - the conflict is arithmetic, not stylistic.
Options with consequences stated: depot test (709 h absorbed off-site, site labour unchanged, adds test cycle time to the critical path); sampling instead of 100% (a variance from R-14 as written, needs client-approved deviation); in-line test at site (breaks R-31 and consumes the site's whole labour budget on testing alone). If depot testing adds a week to the critical path, that is 7 × $4,800 = $33,600 of exposure, which is the price to compare against the cost and schedule of negotiating a deviation. The recommendation should follow the numbers, not the delivery lead's preference.
2. How should the decision be recorded and governed so the baseline is trustworthy afterwards?
Read the worked answer
Raise a formal change request against the requirements baseline that documents: the discovered gap, the capacity analysis, the options considered, the decision and its rationale, the residual risk, and the client's written acceptance. The baseline was signed without this decision, so the correct action is to amend the baseline - not to leave the decision implicit in how the work happens to be executed.
Update the traceability matrix so that R-14 maps to an explicit verification method, location, and evidence artefact (depot test record shipped with each unit), and so that R-31 is annotated with what it does and does not constrain. Record the assumption in the plan in falsifiable terms, so a later reader can test it.
The delivery lead's proposal to absorb it in contingency is the specific thing to refuse. The contingency will be consumed whether or not the analysis holds, no one will be able to reconstruct why, and the schedule will degrade with no attributable cause. Absorbing an unanalysed decision in contingency does not avoid a change request; it defers it past the point where anyone remembers the reasoning. Board the spend against a named risk with a trigger date.
3. What do you do about the three sites that close access after week 18?
Read the worked answer
Separate sequencing from technical scope. Access closure is a sequencing constraint: those three sites should be scheduled first within the depot test cycle, with shipment by roughly week 16-17 so there is buffer before the window closes. That decision can be made immediately and does not require resolving the test-location question.
If the depot test cycle cannot support those three sites inside the window, that is the trigger to open a scoped negotiation with the client - named sites, named dates, named partial-scope or deferred-test arrangements with a defined completion route - rather than to quietly skip test at those sites. The failure mode to avoid is letting an access deadline silently drive a technical decision; once the work is done, no one can tell whether the test was waived deliberately or omitted by drift.
Check dependent work as well: rework capacity at the depot, shipping lead time, and whether units arriving late from those three sites become the constraint on the remaining nine. The early-closing sites should be the first thing planned and the last thing allowed to absorb any slip.