CAPES cheat sheet

A schedule is a project model of activities, logic, durations, and dates; a schedule model plus the reporting/use process constitutes the project schedule.

AACE International Recommended Practice 29R-03 defines schedule classifications by maturity; the CAPES exam references the AACE 29R-03 schedule classification framework for schedule development maturity.

AACE International Recommended Practice 29R-03 defines schedule classifications by maturity; the CAPES exam references the AACE 29R-03 schedule classification framework for schedule development maturity.

Activity: a discrete, measurable unit of work with defined start and finish; an event/milestone has zero duration and marks a point in time.

Mandatory constraint (e.g., Must Finish On, Must Start On) overrides logic and can create negative float; soft constraints (Start No Earlier Than, Finish No Later Than) restrict without breaking logic.

Open end: an activity lacking a predecessor or successor; dangling logic reduces schedule quality and is flagged in DCMA-style checks.

Lags add wait time between activities; leads overlap them. Both are logic links in CPM and must be justified and minimized.

Float (total float) = late finish - early finish = late start - early start; free float = early start of successor - early finish of activity.

The critical path is the longest path through the network; its total float equals zero (or the minimum value if negative float exists).

WBS decomposes scope hierarchically into deliverable-oriented work packages; activities are assigned to the lowest-level work package.

A work package is the lowest WBS level at which cost and schedule can be managed; it should map one-to-one with schedule activities or activity groups.

AACE 29R-03 schedule classification levels: Class 5 (0-2% definition) through Class 1 (50-100% definition); higher class = more mature schedule.

Schedule basis document records assumptions, calendars, constraints, coding, and exclusions used to build the schedule.

Activity coding (phase, area, discipline, responsible party) enables filtering, grouping, and earned value rollups.

Calendars define working days, shifts, and holidays; mismatched calendars between linked activities cause logic gaps or unexpected float.

Rolling wave planning progressively elaborates near-term activities in detail while far-term work remains at summary level.

A milestone schedule (Level 1) shows key dates only; a Level 3 control schedule shows detailed activities and logic for execution.

Finish-to-Start (FS): successor starts after predecessor finishes; most common and preferred relationship.

Start-to-Start (SS): successor starts after predecessor starts; requires a lag to be meaningful.

Finish-to-Finish (FF): successor finishes after predecessor finishes; common for overlapping finish-driven work.

Start-to-Finish (SF): successor finishes after predecessor starts; rare, used for just-in-time handoffs.

Duration estimating methods: analogous (top-down, historical), parametric (unit rate x quantity), three-point (PERT: (O+4M+P)/6), and bottom-up.

PERT standard deviation = (P - O)/6; variance = ((P - O)/6)^2.

Productivity-based duration = quantity / (crew size x productivity rate), adjusted for calendar and weather factors.

Hard logic (mandatory) reflects physical/technical sequence; soft logic (discretionary) reflects preference and can be resequenced.

Forward pass: ES = max(EF of predecessors + lag); EF = ES + duration. Backward pass: LF = min(LS of successors - lag); LS = LF - duration.

Total float = LS - ES = LF - EF. Free float = min(ES of successors) - EF.

Critical path = path with minimum total float (usually zero); multiple critical paths can exist.

Negative float indicates a constraint or imposed date that cannot be met with current logic and durations.

A hammock activity summarizes work between two dates or activities and does not drive logic; its duration is derived, not estimated.

Near-critical path has low total float (commonly <= 10-20% of remaining duration) and should be monitored for risk.

Total float is shared along a path; its consumption by one activity reduces float available to others on that path.

Longest path method (as used with multiple calendars and lags) identifies the driving path more accurately than float alone.

Resource histogram plots demand per period against available capacity; over-allocation appears as peaks above the capacity line.

Resource leveling adjusts start/finish within available float to smooth demand; it may extend the project duration.

Resource smoothing (time-limited) shifts work within float without changing the critical path or end date.

Resource loading assigns quantities and cost rates to activities; loading precedes leveling.

Resource-constrained schedule: logic plus resource limits determine dates; float may be consumed or negative.

Peak resource demand drives cost and risk; leveling targets reduce peaks and idle time.

Crew balance and continuity of work reduce mobilization/demobilization costs on linear and repetitive projects.

Cost-loaded schedule assigns budgeted cost to activities; total cost-loaded value should reconcile to the project budget.

Earned Value: PV = planned % complete x BAC; EV = actual % complete x BAC; AC = actual cost incurred.

SV = EV - PV; CV = EV - AC; SPI = EV/PV; CPI = EV/AC.

EAC = BAC / CPI (typical variance); EAC = AC + (BAC - EV) (atypical).

Schedule-cost integration requires consistent WBS/cost account coding between schedule and cost system.

Time-phased cost (cash flow) curve derives from cost-loaded schedule; S-curve compares planned vs. earned vs. actual.

Cost loading should not distort logic; costs assigned to summary levels cannot be earned at activity level.

Schedule risk analysis (QSRA) uses Monte Carlo simulation of activity duration ranges to produce a distribution of completion dates.

Three-point estimates feed simulation: optimistic, most likely, pessimistic per activity.

P-level dates (e.g., P50, P80) express confidence: P80 means 80% probability of finishing by that date.

Contingency (schedule reserve) = P-level date - deterministic date; held at project level, not baked into activity durations.

Risk drivers: critical/near-critical activities, long-duration activities, open ends, lags, and constraints.

Correlation between activity durations in simulation reflects common-cause risk and increases the spread of results.

Risk register links threats/opportunities to activities for targeted mitigation and analysis.

% complete methods: duration-based, units completed, milestones, weighted steps, and physical measurement; choose per activity type.

Earned schedule: ES = period where PV equals current EV; SPI(t) = ES / AT; SV(t) = ES - AT (time units).

To-complete performance index: TCPI = (BAC - EV)/(BAC - AC) for budget; (BAC - EV)/(EAC - AC) for EAC.

Forecast completion date = data date + (remaining duration / SPI) or via earned schedule projection.

Progress must be measured objectively; subjective % complete inflates EV and distorts forecasts.

Out-of-sequence progress (work started before predecessors finish) must be identified and corrected in updates.

Performance analysis compares planned vs. actual at activity, WBS, and project levels; variances drive corrective action.

Baseline schedule is the approved, time-phased plan against which performance is measured; changes require formal approval.

Change control process: identify, evaluate impact (scope, cost, schedule), approve/reject, document, and re-baseline if approved.

Re-baselining resets the performance measurement baseline; should be rare and formally authorized.

Schedule revisions (updates) record progress without changing the baseline; re-baselining changes the baseline itself.

Change log tracks all schedule changes, rationale, approver, and date.

Impact analysis uses CPM to determine effect on critical path, float, and completion date before approval.

Baseline integrity requires version control and traceability of approved changes.

DCMA 14-point assessment checks logic, leads/lags, hard constraints, high float, negative float, high duration, invalid dates, resources, and missed activities.

Common DCMA thresholds: no more than 5% of activities with lags; no more than 5% with hard constraints; no more than 5% with total float > 44 days (high float).

Open ends: every activity except start/finish should have at least one predecessor and one successor.

Schedule health checks include logic density, critical path continuity, and duration outliers.

Reporting: Gantt, network diagram, resource histogram, S-curve, and earned value reports tailored to audience.

Data date (status date) is the point through which progress is recorded; all actuals must be current to that date.

Validation confirms the schedule reflects the plan, is logically sound, and reconciles to cost and scope.