Near-Miss Trajectory Review¶
Retrospective analysis — instantiates Layered Defense Gap Decorrelation
Reconstructs the path each real near-miss actually took through the layers and treats it as hard evidence that holes are already starting to align.
A tabletop review imagines how a hazard could travel the stack; a Near-Miss Trajectory Review studies how one actually did — and was caught only at the last layer, or by luck. Its defining move is to treat every near-miss not as a happy ending but as a realized partial trajectory: a documented case in which the hazard already passed through several barriers' holes and stopped one short of harm. Each event is reconstructed layer by layer — which holes it slipped through, which barrier finally caught it, how close the catch was — and filed in a repository of real paths. Because these are events that happened, the mechanism reads them as weak signals that specific holes are aligning right now, not hypothetically; and each reconstructed path earns a targeted intervention to close the route it exposed. Where the prospective review works forward from imagination, this one works backward from evidence.
Example¶
An airline's safety team reviews a month of runway-incursion near-misses at one hub. In three separate cases, an aircraft crossed a hold-short line without clearance and was stopped only by a controller's last-second call or the other aircraft's rejected takeoff. Reconstructing each trajectory through the layers, the review finds a common shape: the readback-clearance check passed nominally (crews read back "hold short"), the airport surface markings were faded at one taxiway, the ground-radar conflict alert was set to a delay that fired late, and the controller was working combined positions during a staffing gap. In each near-miss, four holes had already lined up; only the final human catch — itself luck-dependent — prevented a collision.
Filed as three entries in the near-miss repository, the pattern reads as a weak signal that the incursion path is close to complete at this hub under combined-position staffing. The review does not wait for the accident to confirm it: it spawns interventions aimed at the exposed route — repaint the faded markings, shorten the conflict-alert delay, cap position-combining during peak. The events, not a hypothesis, drove the fix.
How it works¶
The review's discipline is reconstruction over blame:
- Treat the near-miss as a full path, not a save. Trace every barrier the hazard already cleared before it stopped, rather than crediting the layer that happened to catch it.
- File the trajectory, build the pattern. Store each reconstructed path in a repository so recurring partial routes become visible across events.
- Read events as drift signals. Interpret repeated near-misses along the same route as evidence that specific holes are aligning now — a live weak signal, not an anomaly.
- Spawn a path-specific intervention. Each confirmed trajectory earns a targeted fix aimed at the route it exposed, closing or staggering one of its aligned holes.
Tuning parameters¶
- Reporting threshold — how minor an event still counts as a near-miss worth reconstructing. Lower thresholds catch earlier drift but flood the repository.
- Reconstruction depth — stop at the proximate save or trace the full multi-layer path. Full paths reveal alignment; shallow reviews relearn single-layer blame.
- Pattern window — how many events, over what period, before a recurring route triggers action. Short windows react fast; long ones separate signal from noise.
- Blame insulation — how strongly reporting is protected from punishment. Stronger protection yields more honest near-miss data and weaker deterrence signaling.
- Intervention trigger — one striking trajectory versus a repeated pattern before a fix is mandated. Single-event triggers are responsive; pattern triggers avoid overreacting.
When it helps, and when it misleads¶
Its strength is that it mines free evidence: near-misses are accidents that already happened minus the harm, and reconstructing their trajectories reveals which holes are aligning before the loss confirms it. Taking those signals seriously is the antidote to normalization of deviance — the slow slide by which repeated close calls get reinterpreted as acceptable rather than as warnings.[n1] Its failure mode is dependence on what gets reported and how honestly: unreported near-misses leave the repository blind, a punitive culture starves it, and hindsight can impose a tidy single trajectory on a messy event. The classic misuse is outcome bias — filing the near-miss as "no harm, no problem" and crediting the lucky final catch as a working defense, which is exactly how the warning is thrown away. The guarding discipline is to protect reporting, reconstruct the whole path rather than the save, and act on recurring routes as if the harm had occurred.
How it implements the components¶
Near-Miss Trajectory Review fills the evidence-and-response components:
near_miss_path_repository— its core artifact: a store of reconstructed real trajectories, each showing the holes a hazard already passed through.latent_condition_monitor— it reads recurring near-miss routes as live weak signals that specific holes are aligning, monitoring drift through events rather than inspections.alignment_breaking_intervention_set— each confirmed trajectory spawns a targeted fix aimed at the exposed route, closing or staggering one of its aligned holes.
It does NOT hypothesize the hazard paths and layers prospectively (hazard_path_inventory, defense_layer_map) — building that forward-looking framing from imagination is the Swiss-Cheese Barrier Review. This review reconstructs the paths hazards actually took; the tabletop review anticipates the ones they might.
Related¶
- Instantiates: Layered Defense Gap Decorrelation — feeds the model the real trajectories that update it after every close call.
- Consumes: Swiss-Cheese Barrier Review for the layer framing its reconstructions map onto.
- Sibling mechanisms: Swiss-Cheese Barrier Review · Barrier Gap Walkthrough · Aligned Gap Heatmap · Bowtie Analysis with Layer Gaps · Common-Cause Layer Audit · Independent Barrier Test Drill · Latent Condition Rounds
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Near-Miss Trajectory Review operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it reconstructs the path each real near-miss actually took through the layers and treats it as hard evidence that holes are already starting to align.
Independent corroboration: The frozen evidence defines Near-Miss Trajectory Review as 'Reconstructs the path each real near-miss actually took through the layers and treats it as hard evidence that holes are already starting to align', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Engineering & Design
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Accident and reliability engineering reconstruct event paths through layered defenses to learn how a near accident traversed real controls.
Related originating lineages:
- Aviation & Aeronautics — Aviation safety institutionalized detailed incident-path reconstruction and defenses-in-depth analysis.
- Medicine & Healthcare — Patient-safety analysis independently adopted layered-defense and Swiss-cheese models.
- Systems Thinking & Cybernetics — Systems analysis supplied attention to interacting layers and emerging alignment rather than a single failed component.
Review resolution: Both independent reviews agree on primary origin engineering_design; reconciliation resolves alternate_origin_disagreement, origin_mode_disagreement. Formative alternate lineages retained: aviation_aeronautics, systems_cybernetics, medicine_healthcare. The broader reach of later applications is kept separate as domain_reach=multi_domain; origin_mode=cross_disciplinary_synthesis describes the historical relationship among lineages. Confidence is conservatively reconciled to high, and encyclopedia_synthesis=true preserves the reviewers' boundary judgment.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
This review is event-triggered — it fires when a near-miss is reported and reconstructs that discrete case. Its sibling Latent Condition Rounds is calendar-triggered, sweeping for drift on a schedule whether or not anything happened; the two share the weak-signal idea but sample it from opposite ends — history versus routine surveillance.
[n1] Normalization of deviance (Diane Vaughan) is the process by which repeated exposure to a deviation with no bad outcome gradually redefines it as acceptable, so that warning signs stop being read as warnings. Near-miss trajectory review exists to counter it, by treating each close call as evidence of aligning holes rather than as proof the system is fine. ↩