Collapsed As-Built, With a Real As-Built Schedule

What Collapsed As-Built Is

Collapsed As-Built (CAB) is a retrospective delay analysis technique. It starts from the schedule’s as-built record -- the actual start and finish dates every activity really carried, recorded after the fact -- rather than from a forward-looking plan. The analyst then identifies one or more specific delay events already reflected in that record, removes their impact from the network, re-links the activities that surrounded the gap the way the work would have flowed without the event, and recalculates the network to get a hypothetical finish date. The difference between that hypothetical date and the real, actual finish date is the delay attributable to the event that was removed.

That’s the opposite direction of a Time Impact Analysis, which is prospective: it inserts a hypothetical delay fragnet into a schedule update before the outcome is known, to forecast the impact of an event as of the date it occurred. CAB instead looks back at a completed (or substantially complete) project and asks what the critical path would have looked like if one specific piece of history had never happened. Because it depends entirely on real actual dates rather than a forecast, CAB is typically performed after substantial completion, when the full as-built record exists to collapse.

Building the As-Built Schedule

Here’s a simplified ten-activity as-built schedule for a small commercial building shell, recorded with the real actual start (AS) and actual finish (AF) dates each activity carried. Activity E carries a delay event: the structural engineer rejected the contractor’s rebar shop drawings over an embed-plate detail, forcing a re-fabrication that paused foundation work for three weeks. For simplicity, this example treats weekends as non-working and observes no holidays.

All activities run Finish-to-Start. Span is the actual calendar-day length of each activity, start through finish inclusive. All dates fall in 2026.

Activity E’s 40-calendar-day span is really three pieces: one week of rebar placement (Mon 30 Mar – Fri 3 Apr), then a three-week suspension (Mon 6 Apr – Fri 24 Apr) while the rejected rebar was re-fabricated to the revised detail, then two more weeks completing the pour (Mon 27 Apr – Fri 8 May). The actual work content is unchanged by the delay event -- three weeks of rebar and concrete work either way -- only the three-week suspension sitting in the middle of it is new.

Collapsing the Delay Event Back Out

To collapse the schedule, the three-week suspension is removed from Activity E and the activity is re-linked directly: the two weeks of pour work that actually followed the suspension are joined straight onto the one week of work that preceded it, exactly as they would have run had the rebar never been rejected. Every activity downstream of E keeps its own actual duration and logic unchanged -- it simply starts three weeks earlier, because its predecessor now finishes three weeks earlier.

The Result

Actual as-built Substantial Completion: Friday, October 2, 2026. Collapsed (hypothetical) Substantial Completion, with the rebar re-fabrication delay removed: Friday, September 11, 2026. The difference is 21 calendar days -- exactly the length of the suspension that was collapsed out, which is what you’d expect on a single unbranched critical chain with no float anywhere in the network: every day added to Activity E’s suspension passed straight through to the finish date, and removing it passes the same 21 days back the other way.

Frequently Asked Questions

Is Collapsed As-Built the same as a Time Impact Analysis?

No -- they sit on opposite sides of a key line. A Time Impact Analysis is prospective: it inserts a hypothetical delay fragnet into a forward-looking schedule update, before the outcome is known, to forecast the delay a single event will cause. Collapsed As-Built is retrospective: it starts from the completed as-built record, after the project is finished (or far enough along to have real actual dates), and removes an event that already happened to see what the finish date would have been without it.

Why is it sometimes criticized as a technique?

The main criticism is that 'collapsing' a network after the fact requires judgment calls about how activities would have been re-sequenced or re-linked absent the delay -- decisions that are speculative because they were never actually tested in the field. Courts and boards that accept the method generally expect the re-linking logic to be well documented and consistent with how the contractor actually would have built the work.

Does Collapsed As-Built require a baseline schedule?

Not necessarily. Unlike techniques that measure delay against a baseline's planned dates, Collapsed As-Built works entirely within the as-built record -- it compares the real finish date to a hypothetical finish date derived from that same as-built network with one or more events removed. A baseline can still be useful for context, but it isn't the basis of the calculation.

Can more than one delay event be removed at once?

Yes. Analysts sometimes collapse out several discrete events in a single pass -- for example, every owner-caused delay -- to isolate the combined effect of one party's responsibility. Each additional event removed adds another re-linking judgment call, which is part of why the technique draws more scrutiny as the number of removed events grows.