CASE 09 · CASE ATLAS · LEVEL B WORKSHOP SCENARIO

Bearing failure learning

SITUATION

A bearing failure has occurred in the field. The design and risk review teams want to draw on every relevant closed investigation and lesson learned before finalizing the corrective design, rather than relying on individual memory of past cases.

EVIDENCE

Closed investigation reports, field-return analyses and lessons-learned records, tagged by component family, failure mode and root cause where recorded.

AI APPROACH

Retrieval surfaces prior investigations and lessons learned that share component family or failure-mode characteristics with the current failure, ranked by relevance for the reviewer to assess.

KEY CONTROLS

Retrieved records are presented with their source and closure status; the tool does not synthesize a combined root cause across records without the reviewer's independent judgement.

VALIDATION MEASURES

Check retrieval recall against a manually compiled list of relevant prior cases for a sample of past failures, so the team knows what fraction of relevant history the tool actually surfaces.

HUMAN DECISION

The design and risk review team decides which retrieved lessons apply and how they change the corrective design or future risk analysis.

DISCUSSION QUESTIONS

  1. What relevant prior case would you be most worried about this tool missing, and why?
  2. How do you weigh a lesson learned from a different component family that only partially matches?
  3. Who updates the tagging on this failure once it closes, so the next search can find it?

TRY IT LIVE

Run this scenario's approach through Quality Evidence Graph — What connects this decision to its evidence, controls, and lessons learned?

Open Quality Evidence Graph →