Nothing quite resembles the mad scramble contractors do to close out the final segments of their projects. By this time, the project schedule has likely been deprecated into a chaotic parallel universe that serves no practical purpose other than to mollify clueless stakeholders. I find this point to average between 40-50% of the overall project duration, but I can see its inevitable coming in the baseline. This is when the implementation of completion schedules can be effective.

As chaos becomes imminent, I advocate the strategy of converting the progress schedule into a completion schedule. The completion schedule is a streamlined version of the forward pass intended to impart a more precise sequence that deprecated progress schedules invariably cannot. The completion schedule will filter out activities that are not impacting completion.

CPM Coordination of Baseline Construction Project Schedules

There are two take-off points when a completion schedule should be launched: when updates no longer represent a clear forward-pass – such as multiple critical or near critical paths, and when invalid schedules are maintained. A completion schedule must be generated anytime stakeholders demand it, thus it is best to beat them to the punch, especially if the working schedule is known to be deficient. Credit must go to the scheduler or contractor savvy and honest enough to realize his schedule no longer serves its intended purpose.

“Some contractor’s schedulers often can’t produce quality to adopt forward logic without wholesale resequencing

Final schedule updates and completion schedules differ in their approach and work product. Schedule updates can easily represent past progress, but are increasingly challenged to represent a coherent and comprehensive forward pass as a project develops. Completion schedules are more practical in that the forward pass dates are derived from boots-on-the-ground  team-consensus, not a desk jockey in thrall to some review agency focusing on technical nuances.

Despite typical approaches, a well manicured completion schedule can be used to both meet contractual requirements and enable and facilitate close out. Here are just a few strategies I adopt when generating my completion schedules:

• Out of sequence: my best practice suggestion is to remove out of sequence relationships upon discovery, or when an out of sequence activity starts – even if it results in activities with open predecessors and successors. If there is no natural successor, it’s fine to tie activities to a collective milestone, such as substantial completion, or even a dummy activity. Once an activity starts, it’s a moot point if its non-driving  predecessors are all deleted. That’s because the team’s success lies in a clear forward-pass, not on resolving past project logic, which can be done any time. Reviewers who are unfamiliar with this strategy might reject a schedule on account of open predecessors. They are over-reliant on DCMA than they are reality – always a losing proposition
• Remaining duration/Expected finish: these columns are the ones I like to use to generate accurate completion dates. Rather than letting activity durations and completion percentages drive the logic, I prefer either remaining duration or expected finish dates taken from the field or people building the project. Remaining duration can be calculated or input. Expected finish is manually input, and does not rely on project logic.
• Suppose an activity is posted with 99% completion for six-months running, and into the final phase. That activity lacks proper project logic and keeps popping up like a bad penny. Rather than try to recreate correct project logic, endeavor to give it an expected completion date: it’s OK to have both early-finish and expected finishes, as well. If you doubt the early-finish date of an activity in the schedule, an inquiry to the site might return more reliable data that you can derive true remaining duration. That’s especially helpful for high-duration activities and others where percentage-complete is driven by a percentage of the duration expired. Sometimes you feel more confident in what the field tells you than what the schedule logic says.
• Lean and mean. Get rid of activities that no longer need tracking. For example, not every submittal and shop drawing needs tracking in the schedule. There should be a shop drawing log for that. Too many schedules I see have every submittal posted, yet never track additional submit/review cycles. Why bother posting an activity if you don’t intend to track it? A completion schedule should only have driving submittal/review activities remaining. A baseline schedule should only track critical path and select driving submittal/review cycles.
• Weekly meetings: projects in their final stages deserve better than 1x/month meetings. Hold a completion meeting each week, and communicate frequently with field personnel to get the best real-time data. In these meetings you will learn more accurate forecasting dates to give to stakeholders who have been asking.
• Filter out completed activities from your completion schedule and show them in another document. They have no bearing on your completion schedule

Finally, due-diligence and adherence to best practices and good housekeeping can maximize the longevity of your baseline, and minimize the effort you need to convert it into a completion schedule. The best schedulers know exactly when it’s time to convert to a completion schedule. They should recognize the need before stakeholders and other non-schedulers do. If not, a grave error will have been made.