Domain 2 is 31% of the PMI-SP exam: the 12 official tasks from WBS to PMB, CPM and float, PERT, resource leveling and a shipbuilding self-test.
Domain 1 decides how the schedule will be built and governed. Domain 2 is where the scheduler builds it: breakdown structures, activities, durations, logic, resources, a risk test and finally an approved baseline. This lesson covers the twelve official tasks, the techniques the exam expects for each, common traps and a short self-test built on a shipbuilding example.
Domain overview
The PMI-SP Examination Content Outline describes Domain 2 as "activities related to defining and sequencing activities and milestones, developing the work and resource breakdown structures, and establishing the performance measurement baseline (PMB)." It carries 31% of the exam, second only to Domain 3 at 35%.
The PMI-SP Handbook gives 170 multiple-choice questions, 150 scored and 20 unscored pretest items, in three and a half hours. At 31%, Domain 2 is roughly 46 to 47 scored questions. The content outline PMI publishes for this credential is dated 2012, so check the PMI-SP page for a newer version before you book.
The twelve tasks are not a strict sequence; in practice you loop between them until the dates work. The figure shows one way to group them. The grouping is ours, the task numbers are PMI's.
Choosing the method, tool, level of detail, calendars and update cycle belongs to Domain 1: Schedule Strategy. Applying those rules to build the model is Domain 2. Once the baseline is approved, the work passes to Domain 3: Schedule Monitoring and Controlling. The PMI-SP exam content overview maps all five domains.
Task 1: WBS, OBS, control accounts and work packages
In the outline: develop the WBS, OBS, control accounts and work packages, through subject matter experts, stakeholders and analysis of the contractual commitments, to ensure completion of the project scope.
The WBS says what is delivered and the OBS says who delivers it. In the PMI Lexicon, a control account is the point where scope, budget, actual cost and schedule are integrated, and a work package is the lowest WBS level at which cost and duration are estimated and managed. Domain 1 explains why control accounts sit where the WBS and OBS cross; here you build that grid.
"Analysis of the contractual commitments" means every contract deliverable lands in a work package. On a newbuild that includes work nobody welds: class and flag approvals, model tests, documentation and delivery papers. See building the newbuild WBS and the WBS and bill of quantities.
Task 2: Define activities and milestones
In the outline: define activities and milestones through subject matter experts, decomposition and the scheduling policies and procedures.
Write each activity as a verb and a noun that make sense alone. "Erect block B12 on berth" can be statused by anyone who walks past the berth; "Hull work" cannot. A milestone marks a significant event, such as "Keel laid", and has zero duration.
For granularity, a widely used yardstick is the DCMA 14-point assessment used by the US Defense Contract Management Agency: it flags incomplete activities longer than 44 working days, about two monthly reporting periods, and expects no more than 5% of activities to exceed that. Long activities are hard to status and hide logic; very short ones make the model costly to maintain.
Rolling wave planning handles distant, uncertain work: plan the near term in detail and the future at a higher level. Hull blocks may be fully detailed in the baseline while accommodation outfitting is still a few summary activities.
Task 3: Estimate activity durations
In the outline: estimate durations using subject matter experts and techniques such as three-point, parametric, analogous and/or PERT.
| Technique | How it works | Shipbuilding illustration |
|---|---|---|
| Analogous | Duration of a similar past activity | The same block on the sister hull |
| Parametric | Algorithm on historical data and parameters | Weld length divided by metres per crew shift |
| Three-point (triangular) | (O + M + P) / 3 | Equal weight to all three |
| PERT (beta) | (O + 4M + P) / 6, standard deviation (P − O) / 6 | Most likely weighted four times |
Illustrative example: the panel line foreman estimates welding a double bottom block at 4 days optimistic, 5 most likely and 9 pessimistic. Triangular gives 18 / 3 = 6.0 days; PERT gives (4 + 20 + 9) / 6 = 5.5 days, standard deviation about 0.83. When the pessimistic value is far out, the two formulas differ, and questions test which one you used.
The GAO Schedule Assessment Guide adds two practices: the people who own the work should estimate it, and durations should reflect most likely conditions without hidden padding. Risk margin belongs in visible contingency. The outline also lists "activity execution techniques (duration/time, effort/work)", the difference between a fixed-length task and one whose length follows from work hours and crew size; in MS Project that is task types and effort-driven scheduling.
Task 4: Sequence activities
In the outline: sequence activities with defined dependencies (internal, external and cross programs), milestones and constraints (for example calendars, geography, contracts), to develop a logical, dynamic schedule model.
"Dynamic" means the logic drives the dates, so the model moves by itself when a duration changes. The Precedence Diagramming Method uses four relationships:
- Finish-to-start (FS): successor starts after predecessor finishes. Welding finishes, then blasting starts.
- Start-to-start (SS): successor cannot start until predecessor starts.
- Finish-to-finish (FF): successor cannot finish until predecessor finishes.
- Start-to-finish (SF): successor cannot finish until predecessor starts. Rare, and a favourite distractor.
A lag delays the successor; a lead advances it. DCMA allows no leads, lags on at most 5% of relationships, and expects at least 90% FS links. The GAO guide adds that lags should not stand in for work or procurement, because a lag cannot be statused.
A common shipbuilding error follows from this: the main engine entered as a milestone with a "start no earlier than" date. The GAO guide prefers modelling the procurement itself (order, manufacture, factory test, shipping, receipt) as activities linked to installation, so a late factory test moves installation automatically. See procurement of major ship equipment.
GAO calls "start/finish no earlier than" soft constraints and "start/finish no later than", "must start on" and "must finish on" hard constraints, because hard ones stop logic moving the activity. DCMA fails a schedule where more than 5% of activities carry hard constraints. For MS Project specifics see task constraints and task dependencies.
Finally, check for incomplete logic: every activity needs a predecessor and a successor except the start and finish milestones. GAO separately defines dangling logic: an activity with links whose start is not driven by any predecessor, or whose finish drives nothing, typical with SS and FF links.
Task 5: Identify critical and near-critical paths
In the outline: identify critical and near-critical paths using techniques such as CPM, Critical Chain, PERT and Monte Carlo simulation, to meet delivery date requirements.
The PMI Lexicon defines the critical path as the longest path through the project, which sets the shortest possible duration. Total float is how long an activity can slip without delaying the project finish; free float, without delaying any successor's early start. Near-critical means total float that expert judgment considers low.
Illustrative network for one hull block (ES = 4 means the activity starts after day 4):
| Activity | Dur. | Pred. | ES | EF | LS | LF | TF | FF |
|---|---|---|---|---|---|---|---|---|
| A Cut and prefabricate plates | 4 | none | 0 | 4 | 0 | 4 | 0 | 0 |
| B Weld panels | 5 | A | 4 | 9 | 4 | 9 | 0 | 0 |
| C Fabricate stiffeners | 3 | A | 4 | 7 | 6 | 9 | 2 | 2 |
| D Assemble block | 6 | B, C | 9 | 15 | 9 | 15 | 0 | 0 |
| E1 Order pipe spools | 5 | none | 0 | 5 | 7 | 12 | 7 | 0 |
| E2 Deliver and inspect spools | 3 | E1 | 5 | 8 | 12 | 15 | 7 | 7 |
| F Pre-outfit pipes in block | 3 | D, E2 | 15 | 18 | 15 | 18 | 0 | 0 |
| G Blast and paint block | 4 | F | 18 | 22 | 18 | 22 | 0 | 0 |
| H Erect block on berth | 2 | G | 22 | 24 | 22 | 24 | 0 | 0 |
The duration is 24 days and the critical path is A, B, D, F, G, H. E1 has 7 days of total float but no free float, because any slip immediately delays E2.
"Critical" and "zero float" are not always the same. The GAO guide warns that multiple calendars, date constraints, out-of-sequence progress and resource leveling distort float, so a constrained schedule can show zero or negative float on activities that do not drive the finish. The longest path is the safer test. Critical Chain, per the Lexicon, places buffers on paths for limited resources and uncertainty. The CPM lesson for shipbuilding goes deeper.
Tasks 6 and 7: Resources and the resource-constrained schedule
In the outline: develop the resource breakdown structure (RBS), determine availability and assign resources (Task 6); adjust the model for resource availability, available budget and other constraints to calculate the resource-constrained schedule (Task 7).
The RBS groups resources by category and type: fitters, welders, pipe fitters, painters, plus cranes, dock and blasting hall. CPM alone assumes unlimited resources; Task 7 removes that assumption.
- Resource leveling adjusts the schedule to optimise allocation and may change the critical path.
- Resource smoothing uses only free and total float, so the critical path does not change.
MS Project's "Level only within available slack" option works like smoothing: it keeps the finish date from moving, and Microsoft notes it may change little when the schedule has not much slack. If two blocks need the same crane in the same week, one moves; if it was critical, delivery moves too. See resource leveling and crane and dock capacity and estimating man-hours.
If leveling pushes the finish too late, compress. Crashing adds resources at the least incremental cost; fast tracking overlaps activities normally done in sequence, adding risk. That is why Task 7 mentions budget.
Tasks 8 and 9: The program plan and the contract dates
Task 8: align the schedule with the overall program plan or integrated master plan (IMP). Task 9: analyse major milestones against the statement of work, contract or memorandum of understanding to see whether the model meets required deadlines.
On a naval program the ship schedule sits inside a larger program; see building a plan for a military shipbuilding program. On a commercial newbuild, instalments are usually tied to events such as steel cutting, keel laying, launching and delivery, explained in how a shipbuilding contract works.
Task 9 is a comparison. If the model shows launching two weeks after the contract date, forcing the date with a "must finish on" constraint hides the problem. The answer is a decision: change the logic, add resources, reduce scope or raise it with the owner.
Task 10: Schedule risk analysis
In the outline: perform schedule risk analysis with quantitative techniques such as what-if scenarios and Monte Carlo simulation, to judge whether milestone dates are achievable within acceptable risk tolerances.
Monte Carlo replaces single durations with ranges, often the three-point estimates from Task 3, and runs the network many times. The output is a probability per date, for example 50% for launching by the contract date and 80% two weeks later. Risk tolerance decides which confidence level to plan for, and the gap becomes visible schedule contingency. GAO lists schedule risk analysis among its ten best practices and discourages hiding margin in lags. See our shipyard risk management article.
Tasks 11 and 12: Consensus, baseline and PMB
Task 11: obtain the consensus of the customer, sponsor, project manager and team to establish an approved baseline schedule. Task 12: establish the performance measurement baseline using organizational processes and standard techniques.
The Lexicon defines the schedule baseline as the approved schedule model, changed only through formal change control and used for comparison with actual results. The PMB integrates the scope, schedule and cost baselines. The time-phased budget turns dates into planned value, the basis of earned value on a shipyard project.
Consensus matters because production managers will not status honestly against dates they never accepted. On a newbuild the "customer" is the owner, so review the baseline with production before it goes to the owner. The newbuild master schedule template is a starting structure.
Typical exam traps
- Total vs free float: the project finish versus the next activity's early start.
- Leveling vs smoothing: leveling may move the critical path; smoothing stays inside float.
- Crashing vs fast tracking: one adds cost, the other adds risk.
- Triangular vs PERT: divide by 3, or weight the most likely by 4 and divide by 6.
- Schedule baseline vs PMB: only the PMB includes the budget.
- Hard constraint as a fix: it hides the problem instead of solving it.
- Domain 1 or 2: choosing the rules is strategy; applying them is Domain 2.
Test yourself
These questions are written for this lesson, not taken from any PMI question bank.
1. In the block network above, E1 slips 4 days. What happens to block erection? A) Slips 4 days. B) No change. C) Slips 3 days. D) E1 becomes critical.
2. Welders are levelled across three blocks without moving delivery. Which technique is this? A) Resource leveling. B) Resource smoothing. C) Crashing. D) Fast tracking.
3. A main engine delivery is a milestone with a "must start on" constraint. Best correction? A) Change it to "start no earlier than". B) Add a 60-day lag from contract signing. C) Model the procurement as linked activities from order to receipt. D) Leave it, the supplier confirmed the date.
4. Estimates are 6, 8 and 16 days. What is the PERT estimate? A) 8. B) 9. C) 10. D) 12.
5. Calculated launch is ten days after the contract date and the project manager asks for a "must finish on" constraint. What first? A) Apply it and baseline. B) Analyse the driving path and present options. C) Add leads to launch predecessors. D) Delete the launch milestone.
6. Which belongs to the PMB but not to the schedule baseline alone? A) Activity dates. B) Time-phased budget. C) Relationships. D) Milestones.
Answers: 1 B, E1 has 7 days of total float (E2 starts 4 days later). 2 B, smoothing stays within float. 3 C, as the GAO guide recommends. 4 B, (6 + 32 + 16) / 6 = 9. 5 B, Task 9 in practice. 6 B, the PMB integrates cost.
Where this leads
Domain 2 ends with an approved baseline. Status updates, variance analysis and change control follow in Domain 3. Study this domain alongside the PMI-SP Examination Content Outline and PMI's Practice Standard for Scheduling, third edition.
