Shared intelligence becomes useful when another team can understand a decision, inspect the evidence and judge whether it applies to their situation. Capture the conditions and uncertainty along with the outcome so a successful investigation can guide later work.
Turn an experience into a usable record
Cookson Electronics used the phrase Shared Intelligence in 2003 for collaboration across its electronics and materials businesses. The enduring problem is practical: valuable experience often remains scattered among supplier emails, test reports and the memories of the people involved.
Start with one resolved or actively bounded problem. Write a task-oriented title, such as “Check packaging history when connector inspection results differ between lots.” Add the affected item family, environment, owner, record revision and review trigger. Use approved internal references for sensitive material and summarize only what the intended audience may access.
NASA’s lessons-learned program describes records that connect a driving event with a lesson and recommendations. The template here uses that broad principle to keep an engineering team’s observations, reasoning and recommended next step together.
Work through a fictional supplier investigation
A change in inspection results
A fictional electronics team notices different inspection outcomes in two incoming connector lots. The supplier and the receiving team compare lot histories and handling records. One lot has a documented packaging change. The initial inspection also used different preparation records, so packaging is a hypothesis rather than an established cause.
The team’s decision is to retain lot traceability, request the missing preparation evidence and plan a controlled comparison under its approved test procedure. This example illustrates evidence handover; it supplies no material settings, acceptance limits or manufacturing instructions.
| Record element | Example entry | Evidence or boundary |
|---|---|---|
| Observation | Lots A and B have different recorded inspection outcomes. | Link report IDs, sample selection and procedure revisions. |
| Known conditions | Packaging differs; preparation records are incomplete. | Link supplier change record and identify the missing record. |
| Hypothesis | Packaging or preparation may explain the difference. | List alternative explanations and the evidence needed to distinguish them. |
| Decision | Preserve traceability and request a controlled comparison. | Name the decision owner and approved test-planning route. |
| Reuse limit | Use this record to structure an investigation. | Apply findings to another item only after reviewing conditions and evidence. |
A colleague should be able to distinguish what was seen from what was inferred without asking the original author. Where a result is unavailable, record “unknown” and the owner of the next inquiry.
Make the evidence package retrievable
Give each supporting report a stable reference, version, date and responsible team. Describe the measurement or observation sufficiently to locate its method, sample and relevant conditions. Link raw results and approved summaries according to access rules. Preserve an explanation of excluded or conflicting evidence.
Capture supplier contributions with their origin and permitted sharing scope. Before moving a record into a wider knowledge library, review confidentiality and remove unnecessary personal details. A useful public-facing summary can point authorized staff to a protected evidence package without exposing it.
For numerical conclusions, retain units, the comparison basis and uncertainty relevant to the decision. For qualitative observations, preserve the exact observation context and explain how the team assessed it. A label such as “improved” needs a stated comparison.
Review the handover with its next reader
- Ask a colleague outside the investigation to find the record using the words they would normally search.
- Have them identify the affected conditions, decision and evidence. Repair missing links or ambiguous labels.
- Ask what would make the result inapplicable. Add those limits beside the recommended action.
- Name an accountable owner and a review trigger, such as a design revision, supplier change or conflicting observation.
- Record who reviewed it and which revision they accepted. Link any unresolved questions to assigned work.
Use the engineering knowledge record and handover checklist (plain text) for a completed investigation or a live record with explicit open questions. It includes separate fields for observations, hypotheses, decisions and reuse conditions.
Revise the lesson when evidence changes
Suppose a later controlled comparison finds that inspection preparation explains the difference while packaging has no observed effect under the tested conditions. Add the new report, revise the conclusion and retain the earlier hypothesis in the revision history. State which earlier recommendation has changed and notify the teams that relied on it through your established workflow.
Track usefulness through concrete reuse: a reader found the record, checked applicability and used or rejected its recommendation with a reason. A large library alone says little about whether engineering experience is reaching the next decision.