Training scenario, illustrative only.Exercise NORTHERN LANTERN is a hypothetical scenario written for training, exercise design and capability demonstration. It does not describe a real operation, a real customer, a real dataset or a live deployment. All organisations, timings and material in it are invented.

The scenario

A fictional five-day exercise brings together a regional police force, a national intelligence customer, a counter-fraud regulator and two authorised partner organisations. The injected threat picture builds around a logistics hub: access-control anomalies nobody can explain, a cluster of suspicious company registrations, haulage manifests that do not reconcile, and a rising volume of open-source chatter.

Every team starts with its own material, in its own format, under its own handling rules. None of them can see the whole picture, and none of them can lawfully give everything to everyone. That tension is the exercise.

The problem the exercise sets

Everything arrives in a different shape: PDFs, scans, spreadsheets, message exports, structured feeds, photographs, with nothing resembling a common schema. The same person, company and vehicle appear under different spellings, registrations and identifiers in each of them. Analysts cannot judge reliability without knowing where a fragment came from and when it was collected. The partners hold different clearances, so one shared workspace is not a lawful answer. And the decision is needed in hours, while collation and re-keying traditionally take days.

Phase 1, Getting information in, with NERVE

Each team connects only the material it is authorised to contribute, and NERVE takes it as it is. Nobody is asked to reformat anything first, which is usually where this kind of exercise stalls.

Documents, exports, feeds and media go into a governed index while the original files stay untouched and retrievable. Each item carries its source, contributing organisation, collection time and handling marking from the moment it lands. Scanned and image-based material becomes searchable alongside native text, so nothing is invisible simply because of its format. The integration is service-mediated: no contributing system is merged, replaced, or given a route into any other.

Phase 2, Resolution and enrichment

Candidate matches across spellings, registrations and identifiers are put to an analyst for confirmation rather than merged silently. Relationships between people, companies, vehicles, locations and events build into a connected picture in which every link can be traced back to the material that justified it. Gaps are stated explicitly, so the team can see what it does not know as clearly as what it does.

Phase 3, Turning information into intelligence, with CORTEX

NERVE gets the information in. CORTEX is where the team turns it into something a decision-maker can act on.

Analysts build assessments against a structured template, recording confidence and source reliability alongside each judgement rather than burying them. Every claim stays linked to the material underneath it, so a reviewer can interrogate the basis instead of taking it on trust. Products are versioned, so when new material changes the picture the assessment is updated and the history survives. The same evidence base produces a short decision brief, a fuller analytical report and a tasking summary, without anyone retyping it three times.

Phase 4, Sharing product and material with authorised partners

This is the part that usually fails in practice. Teams either over-share, or share nothing and lose the advantage. The scenario tests the middle path: share deliberately, at a defined level, with the provenance still attached.

Release is an explicit decision: a named releasing officer, a recorded justification, a defined recipient group. Product and material are separable, so a partner can be given the assessment, the supporting material, or a redacted subset of either. Handling markings travel with the release instead of sitting in a covering email. A partner can see that a judgement rests on corroborated material without being handed the protected source behind it. Every release is logged, so the exercise can be reconstructed afterwards in full: who shared what, with whom, and on what basis.

Exercise timeline

Illustrative outcomes

Collation effort moves from re-keying and reformatting to analysis. Assessments come out in hours and are revised as new material lands rather than rewritten from scratch. Partners receive product they can act on without receiving material they are not cleared to hold. And the post-exercise review is evidential rather than anecdotal, because the decision trail was captured while the exercise was running.

Governance and assurance

Contributing organisations keep ownership of their own material throughout; nothing is absorbed into a pooled holding. Access is governed by role and clearance and enforced at the point of query rather than by convention. Originals are preserved unaltered, so anything relied upon can be re-examined or disclosed later. Ingest, query, production and release are all logged, which gives a defensible record for oversight and, where it matters, for court.

Why we built the exercise this way

Joint working between intelligence and policing rarely breaks down because the analysis is weak. It breaks down because material cannot be brought together quickly, because judgements cannot be traced back to their basis, and because sharing is either too slow or too blunt. NERVE deals with the first. CORTEX deals with the other two. Run together, a joint team can move at the speed of the threat and still leave a record that survives scrutiny.

Exercise NORTHERN LANTERN is a hypothetical training construct. If you would like to run a tailored scenario against your own operating model, we can design and facilitate one with your team.

Explore the platforms used in this scenario: nerve.aic.io and cortex.aic.io.