ISMN 5370 Ex2 S6

Activity list

Master record of each unit of work: ID, verb-noun description, predecessors/successors, resources, and assumptions/constraints; each activity is sized at roughly 8–80 hours of effort

Work package vs. activity

Lowest WBS level describing what is delivered vs. the decomposed unit describing how the work gets done

Finish-to-Start (FS)

B can't begin until A ends; the most common dependency (e.g., testing after coding)

Start-to-Start (SS)

B can't begin until A begins (e.g., documentation once drafting begins)

Finish-to-Finish (FF)

B can't end until A ends (e.g., quality review closes after inspections)

Start-to-Finish (SF)

B can't end until A begins; rare (e.g., night shift ends when day shift starts)

Mandatory dependency

Hard logic; cannot be changed

Discretionary dependency

Soft logic based on best practice; can be changed, and is the only kind that allows overlapping activities

External dependency

Relationship driven by something outside the team's control (e.g., permits); carries risk

Precedence Diagramming Method (PDM)

Network diagram format where activities are nodes; each node shows duration, ES, EF, LS, LF, and float

Analogous duration estimate

Based on similar past projects; fast and cheap but least accurate, so suited to early stages

Parametric duration estimate

Statistical rate applied to quantity (e.g., 100 m of cable = 2 days, so 500 m = 10 days)

Three-point (PERT) estimate

Expected time = (O + 4M + P) ÷ 6; standard deviation = (P − O) ÷ 6

Duration estimate: expert judgment

Use experienced people to validate other estimates or when the work is new

Forward pass

Left-to-right calculation (EF = ES + duration) that yields the earliest project finish

Backward pass

Right-to-left calculation (LS = LF − duration) that yields the latest dates that don't delay the project

Total float

LS − ES (or LF − EF); slack shared along an entire path

Free float

ES of the successor − EF of the activity; slack belonging to one activity alone

Critical path

Longest path through the network; zero slack, so any delay pushes the end date

Crashing

Compressing the schedule by adding resources or overtime; always raises cost; main risk is diminishing returns; do the cheapest cost-per-day first

Fast-tracking

Compressing the schedule by overlapping activities planned in sequence; no direct cost increase; main risk is rework

Which activities get schedule compression?

Critical-path activities only

Milestone

Zero-duration event marking a point in the project (e.g., "Go-Live")

Gantt chart

Most common schedule format; bars plotted on a timeline

Schedule baseline

Approved schedule; altered only via formal change control; basis for SV and SPI