UPES-ECS Drill & Test SOP¶
For: ERT Lead / Incident Commander + IT / UPES-ECS Admin Purpose: test the emergency system safely — without causing panic or polluting real emergency records.
Golden rule: routine testing uses 199, never 111. Real-111 testing happens only in a planned, announced drill window.
1. Test numbers¶
| Number | Use |
|---|---|
| 199 | Drill/Test line. Simulates the full 111 flow. No real dispatch. Logs labelled DRILL-ONLY. Recorded, marked drill. |
| 198 | Echo test — hear your own audio back (checks mic/speaker/network). |
| 111 | Real line. Test only during an approved, announced drill window. |
| 196 | Internal AI test line (later phase). |
- Any authenticated user may call 199 to check their own setup.
- Drill mode (which drills actually exercise escalation/paging/conference) is controlled by the ERT Lead.
2. Drill labelling¶
Every drill call, recording, and log entry is marked so it can never be confused with a real event:
Paging drills additionally start with the spoken prefix:
"Drill, drill, drill. This is a UPES-ECS drill. No real emergency response will be dispatched."
3. Drill types & cadence¶
| Drill | Frequency | What it proves |
|---|---|---|
| 1. Technical test | As needed / daily readiness | SIP registration, 199 routing, ERT answer, recording, voicemail, logs |
| 2. ERT call-handling drill | Monthly | Operators answer, ask the 6 questions, classify, dispatch, log |
| 3. Missed-call drill | Monthly | Voicemail records + Missed Emergency Review Queue works |
| 4. Paging drill | Monthly (prior notice required) | Authorized paging works; unauthorized paging is blocked; audibility per zone |
| 5. Full incident drill | Quarterly (ERT Lead + university approval) | End-to-end: 111 → queue → dispatch → 9000 → paging → recording → review |
4. Daily readiness check (before relying on the system)¶
Owner: IT/Admin duty person or ERT control-room assignee.
- [ ] Asterisk / FreePBX service running.
- [ ] ERT queue has ≥ 2 available responders.
- [ ] ERT answering device registered.
- [ ] 199 test call works end to end.
- [ ] Emergency recording works (test call produced a file).
- [ ] Emergency voicemail can record.
- [ ] Storage not full (warning at 75%, critical at 90%).
- [ ] Critical fixed phones (Medical 4200, Security 4300) registered.
- [ ] One mobile SIP test over Wi-Fi works.
Log result as Daily ECS Readiness Check.
5. Weekly drill health check¶
Owner: ERT Lead + IT owner. Run a small scenario and verify:
- [ ] Student mobile calls 199 → ERT answers.
- [ ] Recording verified + labelled drill.
- [ ] Missed-call voicemail path tested.
- [ ] Warm transfer tested.
- [ ] Three-way bridge tested.
- [ ] Conference 9000 join tested (authorized) + blocked (unauthorized).
- [ ] Paging tested from an authorized device + blocked from a student account.
Log result as Weekly ECS Drill Health Report.
6. Running a full incident drill (quarterly)¶
Before: 1. Get ERT Lead + university authority approval. 2. Announce the window to ERT / control room (and campus if paging is involved). 3. Take a config backup. 4. Confirm daily readiness check passes.
During: run the scenario end to end using 199 where possible. If the real 111 line must be exercised, announce it first and keep the window tight.
After: produce the UPES-ECS Post-Drill Review Report:
Scenario:
Participants:
Timings (answer / dispatch / resolution):
Pass / Fail per step:
Issues found:
Action items → Owner → Due date:
- SOP issues → action-item owner is the ERT Lead.
- Technical issues → action-item owner is the IT Admin.
- Drill records are kept. Every failure creates an action item.
7. Safety guardrails (what a drill must NOT do)¶
- ❌ Never surprise-test campus-wide paging (700) — always give notice.
- ❌ Never use 111 casually for routine testing — use 199.
- ❌ Never auto-open Conference 9000 for every drill call.
- ❌ Never dispatch a real response during a drill unless the drill controller explicitly approves it.
- ❌ Never leave a drill's simulated "missed" call unreviewed — treat the workflow as real.
8. Go / no-go rule¶
If a 199 or 111 test fails, the system is not ready. Do not go live and do not rely on it until the failure is fixed and re-tested. A failed recording or a queue with zero available responders is a critical failure, not a warning.