The result
Keep the compared versions, computed movement and unexplained portion visible.
A later finish is the start of the question.
An update arrives. The forecast finish has moved. Before you can explain it to someone else, you need to know what changed, what the analysis supports, and what evidence is still missing.
Follow one review of the generated bridge-widening example. Every figure below comes from the published sample results. This is an illustrative workflow, not a customer success story.
Already know what you need? Open the review workspace or explore the schedule directly. No signup is needed for the public example.
The finish moved. Start with the comparison.
Compare the baseline with Update 01. The engine's forecast finish moves from 29 May to 10 June 2026: 8 working days on calendar C1, or 12 elapsed calendar days. All 18 activities match by activity code.
Baseline29 May2026 · computed early finish
Update 0110 June2026 · computed early finish
Your next question: How much of that movement does the analysis actually explain?
Inspect this exact comparison ↗
These are recorded day-based engine calculations. The schedule viewer shows exported dates, which can differ from computed dates. This sample is separate from the newer minute-mode benchmark.
One day attributed. Seven still unexplained.
The comparison attributes a net 1 working day to a named activity and leaves 7 working days as residual movement. Adding those numbers gives the finish movement; it does not prove that every day has an established cause.
1 day · attributed by the model7 days · residual / unexplained
Does that mean one party caused the delay?
No. The comparison identifies a modeled contribution, not contractual responsibility. The remaining movement needs investigation; a schedule difference alone does not establish entitlement or an opinion.
The named contribution is associated with EXCAVATE. Inspect its dates and connected work before relying on the explanation. The viewer will open Update 01 with that activity selected.
Follow the activity and its dependencies ↗
Your next question: What information would let the review move forward?
Turn “not enough information” into a specific request.
In the complete sample's clause review, 68 clauses could not be decided. The missing-evidence report groups their needs into 43 distinct inputs. Delay-event evidence for the ASCE checks is the sole missing input for 12 clauses.
This is a separate view of the complete sample, not proof of what caused the eight-day movement. It helps you ask a precise question instead of sending a general request for more documents.
- Assign a reviewer. Name who will assess the response.
- Record the request. Describe the evidence needed and the source you expect.
- Keep the response with the request. Attach the supporting file locally and record what you reviewed.
Will supplying that input make the 12 clauses pass?
No. It may make them decidable. A reviewer still needs to assess the evidence and rerun the applicable analysis. Marking a request “Reviewed” in the workspace does not change an engine finding.
Notes and attachments stay in the browser workspace. They are not uploaded. Export a package before changing devices or clearing site data.
Leave the next reviewer a starting point.
A useful handoff says what changed, what remains unresolved, which evidence was reviewed and what should happen next. The workspace can export the recorded comparisons, request definitions, your notes and attachment bytes together.
The evidence
Preserve the request, owner, reviewer notes and attachment hashes.
The decision
For a pilot, agree the acceptance criterion, name the reviewer and record the observed result.
Prepare a review handoff ↗ Inspect the engine's reproduction record
The personal workspace is not a shared audit system. A reviewer-entered status is not a professional approval, independent timestamp or certification.
Bring the workflow to your organization
Start with a pilot question, not a promise that every schedule will produce the same result. Which decisions must the report support? Who will review exceptions? What evidence and output formats do they need?
- Agency or DOT review: agree the governing requirements, submission evidence and acceptance authority.
- Contractor review: agree how updates, progress records and evidence requests will move between the people doing the work.
- Forensic review: agree the input validation, method assumptions and reproduction checks another reviewer will need.
Define your pilot acceptance checks ↗ Discuss your workflow with Charley
Check the underlying evidence
The comparison comes from the six recorded engine comparisons, specifically baseline to Update 01. The request counts come from the complete missing-evidence report. Read the input qualifications, limitations and current measurement scope alongside them.