# A2 — Restore the Relay: campus trial

Pairs · 25% · checkpoint 23 April; final 30 April 2027, 17:00 Canberra time.

## Challenge and information split

Restore a silent relay and establish that the correct record reached the receiving desk. In-person is preferred. Divide these cards between two students; do not give one person both responsibilities for the whole run. Swap after the checkpoint or during the revised run.

**Sender — Systems/Investigator:** owns the equipment controls, original measurements, source and copy records. Brings the independent Lab 5 diagnosis trace, verifies the supplied relay before replacement and justifies integrity/permission.

**Receiver — Coordinator:** owns the destination, acceptance conditions and acknowledgement code. Must question incomplete messages, read back the actual profile and identify a stale instruction.

Observer is an optional audit responsibility. Solo practice may switch cards openly; the assessed task remains paired unless an arrangement is recorded.

## Sender card

Supply: 6 V through fuse, switch, load and return. The load is dark. This compact relay supplies a combined source-and-fuse check. Predict its result, record the finding before replacement and verify the load afterwards. Independent selection among discriminating measurements is established in the attached Week 5 record, not by this single supplied check. Isolate before replacement. Ask the facilitator for a requested test result; use the separate [assessment facilitator sheet](assessment-facilitator.md).

Archive A: complete sequence, content reference M27-A, recorded 09:10.
Archive B: complete sequence, content reference M27-B, recorded 09:25.
Original signed source: M27-A. Recency alone does not establish authority.

Permission mandate: Observer/read, Technician/service, Registrar/certify.
Installed policy permits these three and Observer/certify. Derive the justified change; check legitimate actions remain usable.

Send the verified profile and supporting evidence. Ask the receiver for its acceptance card rather than inventing a destination or assuming acknowledgement.

## Receiver card

Destination: Dispatch.
Accepted item: verified archive.
Required condition: after integrity check.
Original receiving code: RELAY-A-FIRST.

Read back the item, destination, condition, verified profile and current code. The sender confirms or corrects this message. Preserve any misunderstanding, including its correction. Agreement by nodding is not a precise read-back.

## Checkpoint and changed requirement

Preserve the original fault evidence, repair verification, source comparison, policy expectations and read-back. At the checkpoint label each person's actions.

For the revised run swap responsibilities. The new signed source is M27-B; the original source remains in the history. The receiving card now says RELAY-B-SECOND. The destination and integrity condition remain the same.

Identify what changes and what can remain. A stale acknowledgement or previously correct copy is insufficient. The equipment-failure preset does not by itself establish an archive change; choose the explicit changed-source case for this task.

## Perform and verify

Once the diagnosed supply, justified source, policy and precise handoff support the transmission, operate the relay terminal. The facilitator provides a receiving acknowledgement. Collect it, check the receiving terminal and deliver the acknowledgement to the receiver.

A working lamp is not proof of receipt; a receipt for the wrong source is not correct delivery.

## Submission

Attach the Week 5 independent diagnosis record: competing causes, chosen test, original readings, repair and verification under a changed condition. Keep original and revised relay records, both students' messages, the full nine-cell permission audit from Week 7 and changed-card communication evidence from Week 8. Submit the 500–700-word joint explanation, responsibility table and individual contribution statements required by the brief.

Record source authority, remaining uncertainty, a consequential misunderstanding and how you corrected it. The model checks conditions; the tutor judges reasoning and collaboration.