An outdated planning basis weakens adherence checks, analysis and forecasting. When the schedule trails the field by a month or more, the picture a director needs has to be rebuilt through meetings and report reviews.
The deck below sets out a standard for a credible planning basis and a way to test it. The reporting lag it shows is illustrative; measuring the real source freshness is part of the pilot.
Use the arrow keys or click to move through the deck. Open it in a full page ↗ · Watch the video on YouTube ↗
Five requirements
The proposed method has to satisfy all five.
- Current project reality. Reflect the current situation across every project phase.
- Links to project plans. Connect the schedule to the Capex, execution, AWP and other relevant plans.
- Complete scope. Represent the full project scope in the schedule.
- Update and change method. Define robust, proven update and change methods from project inception.
- Validated adherence. Check that the schedule remains consistent with the other project plans.
Requirements 2 and 5 are distinct: a link establishes traceability, while an adherence check tests consistency.
A cycle, not a document
Scope, plan links, owners, update rules and change control are defined at project inception. Then the cycle repeats:
- Capture field evidence and validate status across project phases.
- Update the schedule to reflect actuals and remaining work.
- Control changes to scope and plans.
- Validate adherence against the current project plans.
- Review and act: assign actions and issue the controlled update.
The pilot: from the portfolio to the evidence
The proposed pilot follows each exception from the portfolio to the project, its schedule versions, the phase or WBS, and finally the activity and its source evidence, with an accountable owner for each finding. The dashboards shown in the deck run on synthetic demonstration data.
Proposed, and not yet an implemented feature: a rolling review window of one month back and three months ahead of the schedule Data Date, with a difference filter against an agreed reference.
How the evidence is handled
There are two input routes: a P6 SQLite .db and P6 XER exports. Dashboard processing is local and read-only, and the source schedule file is never uploaded. The built-in guide in the .db pilot needs no AI account or key. An AI Analyst is a separate, optional step.
The worksheets and dashboards in the deck are illustrative and use synthetic demonstration data; they are not customer findings.