Dependency Network and Critical-Path Map¶
Analysis — instantiates Temporal Orchestration Design
Maps precedence, float, and alternative paths so the chain of activities that actually governs the finish date — and the slack that does not — becomes visible.
A list of tasks with durations tells you nothing about which ones matter for when you finish. Dependency Network and Critical-Path Map builds the directed graph of what must precede what, threads durations through it, and finds the longest chain of dependent activities — the critical path, whose length sets the earliest possible finish and whose every member has zero slack. Its defining move is separation: it distinguishes the activities that govern the end date from the far larger set that carry float (they can slip without moving the finish) and shows which chains could run in parallel versus which are forced into sequence. It is a diagnostic map, not a decision — it says where a delay would propagate to the deadline, not what protective slack to spend against that risk.
Example¶
An authority is building a new airport terminal and has roughly 900 activities on the plan, from pile-driving to jet-bridge commissioning. Everything has a duration; nobody can say which delays would actually push back the opening. The team assembles the dependency network: structural steel cannot start until foundations cure; the baggage-handling system cannot be tested until its hall is weather-tight; security lanes cannot be certified until power and data are live. Threading estimated durations through the graph, the longest dependent chain runs foundations → superstructure → roof/envelope → baggage-hall fit-out → integrated systems test. That chain — roughly 26 months of the plan — is critical: a week lost anywhere on it is a week lost on the opening. Landscaping and retail fit-out, by contrast, show 4–5 months of float and can absorb slips without consequence.
The output is not a schedule but a diagnosis: these fifteen activities govern the opening date; these six hundred do not. That single distinction redirects management attention — daily standups now track the envelope-and-systems chain, and a proposal to accelerate landscaping is dropped because its float means it was never the constraint.
How it works¶
- Build the precedence graph. Encode each dependency as a typed edge (finish-to-start, start-to-start, with any lead or lag), and flag which orderings are hard prerequisites versus reversible preferences.
- Forward and backward pass. Sweep forward to get each activity's earliest start/finish, then backward from the target finish to get latest start/finish; the gap between them is that activity's float.
- Extract the critical path. The connected chain of zero-float activities is the path that sets the finish; identify near-critical paths (small float) as the next things to watch.
- Read parallelism and contention. Chains with no dependency between them may run concurrently; where two such chains draw on the same crew or crane, mark the resource contention that the pure logic hides.
Tuning parameters¶
- Network granularity — how finely activities are decomposed. Finer resolution locates the constraint precisely but multiplies maintenance; coarse networks hide the true critical chain inside summary bars.
- Dependency strictness — how many edges are modeled as mandatory versus discretionary. Over-linking manufactures a rigid path that isn't real; under-linking lets the map bless impossible overlaps.
- Duration point vs. range — a single duration per activity or a distribution. Ranges reveal that the "critical" path is only probably critical, at the cost of a heavier model.
- Resource overlay — whether the map is pure logic or resource-loaded. Overlaying crews and equipment exposes contention but can shift the critical path to a resource-critical chain.
- Recompute trigger — how large an actual-vs-plan deviation forces a re-pass, so the identified path stays current as work lands.
When it helps, and when it misleads¶
Its strength is focus: on a plan of hundreds of activities it names the handful that actually decide the finish, so acceleration effort, buffers, and management attention all land where they change the outcome instead of being spread evenly over work that has slack to spare. The method is old and well understood.[n1]
Its failure mode is trusting the topology more than the durations feeding it. The critical path is only as real as its estimates: with wide uncertainty, several paths are near-critical and the nominal longest chain can be overtaken by a slip elsewhere, so a team that hardens the "critical" path while ignoring a near-critical one is guarding the wrong chain. The classic misuse is treating a pure-logic map as executable while ignoring that two parallel chains need the same tower crane — the network says they overlap, the yard says they cannot. The guarding discipline is to carry duration uncertainty into the analysis (watch near-critical paths, not just the longest), overlay real resource contention before believing the parallelism, and re-run the passes as actuals arrive rather than freezing the path drawn at kickoff.
How it implements the components¶
precedence_dependency_and_reversibility_map— its core artifact: the typed graph of what must precede what, with mandatory versus reversible orderings flagged.duration_effort_and_elapsed_time_model— it threads activity durations through the graph to compute earliest/latest times and float; elapsed estimates are what make a path "critical."parallelism_concurrency_and_contention_policy— independent zero-dependency chains are precisely the work that may proceed concurrently, and the resource overlay marks where that concurrency contends.
It does not implement buffer_slack_and_variability_absorption — computing where float exists is diagnosis; deciding how much protective slack to place against it is Buffer and Float Allocation Model, the sibling that consumes this map.
Related¶
- Instantiates: Temporal Orchestration Design — supplies the dependency-and-float backbone the rest of the orchestration reasons over.
- Sibling mechanisms: Buffer and Float Allocation Model · Rolling-Wave Schedule · Event-Triggered Rescheduling Controller · Phase-Alignment and Staggering Plan · Time-Window and Readiness Gate · Timebox and Timeout Rule · Cadence Calendar with Adaptation Rule · Temporal Scenario and Stress Test · Cycle-Time and Lag Retrospective
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Dependency Network and Critical-Path Map operates as a computation, comparison, model, or analytic representation used to infer, estimate, or choose because it maps precedence, float, and alternative paths so the chain of activities that actually governs the finish date — and the slack that does not — becomes visible.
Independent corroboration: The frozen evidence defines Dependency Network and Critical-Path Map as 'Maps precedence, float, and alternative paths so the chain of activities that actually governs the finish date — and the slack that does not — becomes visible', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Single lineage
Present-day reach: Multi-domain
Rationale: Industrial project scheduling cohered CPM and PERT network maps that identify precedence, float, and the longest chain governing completion time.
Related originating lineages:
- Organizational & Management Science — Project management institutionalized critical-path maps as planning and coordination artifacts.
Review resolution: Industrial project scheduling cohered CPM and PERT network maps that identify precedence, float, and the longest chain governing completion time. The retained alternate lineages materially shaped the mechanism's form.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] The Critical Path Method, developed in the late 1950s for industrial project scheduling, computes the longest dependent chain through an activity network and the float on every other activity; its probabilistic cousin PERT models durations as distributions rather than points. ↩