Your Gantt Chart Has Blind Spots. Topology Finds Them.
The tasks most likely to fail your project don't appear on the critical path. IRDME's Project Risk Topology Report finds them by comparing declared schedule dependencies against resource sharing and risk coupling.
Every project manager knows the Gantt chart. Boxes, arrows, critical path highlighted in red. The logic is simple: a task on the critical path delays everything downstream. Optimize those, and the project stays on track.
This is a useful model. It is also incomplete in a specific, measurable way.
The Gantt chart is a single-layer graph. It encodes one type of relationship: declared schedule dependency. Task A must finish before Task B can start. That is d1 — the declared structure.
What it does not model is what happens at d2 and d3.
d2 — resource sharing: Which tasks compete for the same people, budget, or physical systems? A task invisible to the critical path can silently starve three others if it consumes the only available structural engineer for six weeks.
d3 — risk co-occurrence: Which tasks, when one fails, tend to cascade failures to others regardless of declared dependencies? This is the layer that matters most in post-mortems — and it appears nowhere in standard project software.
The gap between d1 and d2/d3 is where projects fail.
The IRDME Approach
- IRDME treats a project plan as a multilayer graph. Each subsystem or task is a node. Three layers of relationships are defined:
dependency_chain(d1) — the Gantt: what must complete before whatresource_sharing(d2) — which tasks share budget, personnel, or physical systemsrisk_co_occurrence(d3) — which tasks co-fail: if one fails, what else fails with it
Then IRDME computes hub importance scores for every node in every layer, and measures the cross-layer Pearson correlation r(d1↔d2) and r(d1↔d3).
A high-ranked node in d3 that is low-ranked in d1 is a hub shadow. The critical path missed it. Risk did not.
The Mars Mission Example
We modeled a 12-subsystem Mars mission architecture: launch_vehicle, life_support, comm_relay, landing_system, propulsion, power_systems, navigation, crew_habitat, science_instruments, surface_mobility, ascent_vehicle, mission_control.
The critical path (dependency_chain) puts launch_vehicle and propulsion at the top. Standard project planning agrees.
IRDME found two hub shadows.
comm_relay ranks low in dependency_chain — it is not a schedule bottleneck. No other task formally waits for it in the Gantt. Yet it ranks as the top hub in risk_co_occurrence with six risk edges, including weight=1.0 connections to navigation and mission_control. Loss of deep space communications means guidance and navigation control go blind. Mission abort. The Gantt marks comm_relay as background infrastructure. Structural risk analysis marks it as the single most dangerous failure point.
power_systems is similarly invisible to the schedule. But it holds weight=1.0 risk edges to life_support and distributes budget to seven other subsystems simultaneously. In the Gantt it appears as ongoing infrastructure. In the multilayer model it is a universal hub in both resource and risk layers — the kind of node whose failure cascades mission-wide.
The r(dependency_chain ↔ risk_co_occurrence) for this dataset is weak — meaning the two layers encode structurally different rankings. That divergence is the finding. It means your critical path and your risk model are not measuring the same things.
What This Means for Real Projects
This pattern generalizes. Software projects: a service might not be on the critical path but every team deploys through it — a resource hub invisible to the Jira board. Construction: a specialist supplier not on the critical path but whose delay cascades to six downstream packages. Drug development: a preclinical assay not on the timeline but whose results gate three parallel Phase I decisions.
In each case, the structural diagnosis is identical: high rank in behavioral coupling (d2/d3), low rank in declared dependency (d1). Hub shadow.
IRDME finds these before they become post-mortem entries.
The Report
- The Project Risk Topology Report generates a full structural audit of any project plan you define:
- Hub shadows (Gantt-invisible, risk-critical) — the failures your schedule does not model
- Budget bottlenecks (chameleons in resource layer) — tasks that appear structurally mid-tier but consume disproportionate shared resources
- Mission-critical systems (universal hubs across all layers) — nodes that rank at the top regardless of which layer you look through
- Cross-layer r values with p-values and confidence intervals — the quantified divergence between your schedule model and your risk reality
The Mars Mission dataset is pre-loaded as a demo. Define your own system as JSON or use the calculator to build nodes and edges interactively.
The Gantt chart is not wrong. It is one layer. The project lives in three.
IRDME is a multilayer structural analysis engine. The Functional Proximity Law — confirmed across 16 domains from proteins to software to finance — states that hub importance scores preserve more strongly between layers encoding functionally similar relationships. In project management: schedule dependencies and risk dependencies are not the same relationship type. The divergence between them is measurable, pre-registerable, and independently verifiable.