What a controls function should produce, and what it should be measured on.
1.0 Planning and scheduling
A critical path built to be used rather than filed. That means logic that reflects the actual construction sequence, durations that come from productivity assumptions somebody is willing to defend, and a level of detail matched to the phase — not a 12,000-line network at concept stage because the software allowed it.
Deliverables: baseline schedule with a documented basis, critical path and near-critical path analysis, progress measurement system, recovery and acceleration options with their cost attached.
2.0 Cost engineering
Estimate through to forecast, with the trail intact from one to the other. Every estimate carries a maturity class and that class is stated wherever the number appears. Change is tracked against the baseline continuously rather than reconciled at period end.
Deliverables: cost breakdown structure aligned to the WBS, estimate basis and class, cost and commitment reporting, forecast at completion with variance drivers named.
3.0 Risk analysis
Quantified rather than assumed, so that contingency has a number behind it and can be defended in front of a board. Schedule risk analysis run against the actual network, not a summary of it, with the correlation assumptions written down.
Deliverables: risk register with owners, quantitative schedule and cost risk analysis, contingency drawdown plan, scenario cases for the two or three uncertainties that actually move the outcome.
4.0 Delivery assurance
Earned value reporting that shows the slip early enough to do something about it, and a reporting pack that a project director can read in ten minutes and act on.
Deliverables: period reporting with trend rather than position, early-warning indicators with thresholds agreed in advance, independent review of schedule and estimate quality, readiness assessment ahead of major stage gates.
How engagements typically start
Most begin one of three ways: a baseline that needs building before sanction, a programme that has lost confidence in its own forecast, or an independent review ahead of a funding decision. All three start the same way — with an honest assessment of what the current data can and cannot support.