Project scheduling software and the critical path method
Almost every serious project schedule is a critical path method (CPM) network: activities with durations, linked by logic, calculated forward and backward to find the earliest and latest dates and the chain of work that sets the finish. Here is how the calculation works, what float tells you, and what to look for in scheduling software.
The network
A CPM schedule has activities, each with a duration in working time on a calendar, and relationships between them. A relationship can be finish-to-start (the most common), start-to-start, finish-to-finish or start-to-finish, and can carry a lag. Milestones are zero-duration events. The work breakdown structure (WBS) groups the activities; it organizes the schedule but does not drive the dates. Logic does.
Forward and backward passes
The forward pass starts at the data date and works through the network: each activity's early start is the latest date its predecessors allow, and its early finish follows from its duration on its calendar. The latest early finish is the project's forecast finish. The backward pass starts from that finish (or a required finish date) and works back: each activity's late finish is the earliest date its successors need, and its late start follows from its duration.
Float and the critical path
Total float is how far an activity can slip without moving the project finish: late dates minus early dates. Free float is how far it can slip without delaying any successor. The critical path is the longest path through the network, the activities with the least total float (zero, or negative where a required date is already missed). Negative float is a warning, not a calculation error: the plan as logically linked cannot meet the date imposed on it.
Resources and leveling
A schedule that ignores resources can show five crews working in a space that holds one. Resource-loaded schedules assign labour and equipment to activities; leveling then delays work, by priority and float, so no resource exceeds its limit. Leveled dates are longer and more honest; the critical path becomes a resource-constrained path.
Progress and the data date
As work happens, actual starts and finishes, remaining durations and percent complete are recorded up to the data date. Everything before it is history; everything after is forecast. A baseline kept at approval is what progress is measured against, through variances and earned value.
Choosing scheduling software
- Does its critical path agree with the tools your contractors use, on the same file?
- Does it read and write the exchange formats you receive: XER and Project XML?
- Can it check schedule quality: open ends, hard constraints, long lags, out-of-sequence progress?
- Does it carry resources, costs and baselines, so earned value and forecasts come from the same model?
- Can it quantify uncertainty with schedule risk analysis?
- Can the people who decide read it, without a licence each?
Questions
Is the critical path always the longest path?
In a network without constraints, calendars that differ or resource leveling, yes. With them, the activities with the lowest total float are a better test, which is why software reports float rather than path length.
What is a good amount of float?
There is no single number. What matters is how float is distributed, whether near-critical paths are close behind the critical one, and whether float is consumed faster than time passes.
Primavera and P6 are trademarks of Oracle Corporation; Microsoft Project is a trademark of Microsoft Corporation; PMBOK is a trademark of the Project Management Institute; AACE is a trademark of AACE International; Last Planner is a registered trademark of the Lean Construction Institute. DeltaSys is not affiliated with or endorsed by them. Standards are described in our own words; see the publications themselves for their text.