A security camera that failed in March gets discovered in July, when someone asks for footage that was never recorded. That is the defining problem with CCTV equipment: it does not go missing, it goes silent. The cameras hang in plain sight, the recorder hums in a cabinet, and the first real test of the system is the incident it was bought for. This guide covers a register that catches failures before they matter - inventorying the full recording chain, recording the details that let you fix faults fast, and building a verification routine that does not rely on luck.
What you will learn
- The dead-camera problem
- Inventory the whole recording chain
- What to record for each device
- Labelling cameras, recorders and switches
- A verification routine for silent failures
- Tools that make this easier
- FAQ
The dead-camera problem
CCTV equipment has a failure profile most assets do not share:
- Nothing checks the output daily. A laptop that dies gets reported within the hour; a camera that dies produces an absence nobody is watching for.
- Systems grow in layers. The original installer fitted eight cameras, a second contractor added four, someone in-house swapped a dead unit for a different model. No single as-built record survives that history.
- The chain has invisible links. A failed PoE switch or a recorder with a full disk kills footage from cameras that are perfectly healthy - and those boxes live in cabinets nobody opens.
- Ownership is split. Cameras belong partly to security, partly to facilities, partly to IT. Equipment owned by three teams gets verified by none of them.
The fix for all four is the same: a register that lists every device in the chain, and a routine that confirms each one is doing its job.
Inventory the whole recording chain
Footage depends on a chain - camera, cable run, switch, recorder, storage. Inventory all of it, not just the parts on the walls: the cameras themselves, the NVRs or DVRs they feed, the PoE switches between them, and any backup power keeping the cabinet alive.
The single most valuable artefact is the channel map: which camera feeds which recorder on which channel. When someone needs footage of the loading bay, the question is never “do we have cameras” - it is “which camera covers that door, and where does it record to”. A register that answers this in seconds is worth more than any individual field in it.
What to record for each device
| Field | Why it matters |
|---|---|
| Asset ID | The handle for a camera nobody can physically reach |
| Mounting location | ”North entrance, above the door” answers the question that matters: what does it cover |
| Make and model | Mounts, spares and replacement choices differ per model line |
| Serial number | Warranty claims, plus insurance and police reports if equipment is stolen |
| NVR and channel number | Maps a footage request to a recorder and a port in seconds |
| Install date and warranty end | Cameras fail with age - this dates the fleet and prices the fix |
| Last verified date | The register’s honesty field: when did a human last confirm this unit records |
| Condition and service notes | Water ingress, repositioned views and past faults stay with the device |
Labelling cameras, recorders and switches
Cameras are the one asset you usually cannot walk up to, so labelling works differently:
- Label the camera before it goes up. A QR label on the body costs nothing at install and saves a ladder later.
- Label the reachable end of the run. Put the asset ID at the patch panel or recorder port each camera feeds - that is where day-to-day identification actually happens.
- Label the cabinet contents. NVRs, switches and power supplies each get their own ID; “the black box” is not a record.
- Mark both ends of new cable runs whenever the system is extended, while the installer still knows which is which.
Tip: when a contractor extends the system, make the handover conditional on serials, channel mappings and label IDs for every new device. Chasing this a year later means tracing cables through ceiling voids.
A verification routine for silent failures
The register tells you what should be working; only a routine tells you what is. A workable cycle for most sites:
- Walk the cameras monthly or quarterly. Live view per camera, lens clean, view unobstructed - growth in front of a lens is a classic summer failure.
- Check the recorder, not just the cameras. Confirm footage retention matches your data retention policy, and that the disk is not silently overwriting sooner than intended.
- Raise a ticket per failure. A dead camera spotted on a walk-round and fixed within the week is a maintenance item; the same camera discovered after an incident is an explanation to write.
- Stamp the last-verified date. Maintenance teams fold the camera check into existing site rounds - it does not need to be a separate job, it needs to be someone’s job.
Tools that make this easier
Most CCTV “registers” start life as the installer’s handover spreadsheet and a PDF site plan. Both are accurate for exactly as long as the system stays untouched - the first extension, swap or repair turns them into historical documents, and verification dates were never tracked anywhere at all.
An asset management tool like AMPthilly gives every camera, recorder and switch its own profile - location, serial number, supplier, purchase and warranty dates, with the installer’s handover attached as a document. A scan of the QR label opens the device’s record in a phone browser, where a fault can be reported with photos; tickets move through statuses from in review to resolved and stay on the device’s history permanently, so the register doubles as the maintenance log. The service desk module arrives with the Starter plan; the free plan covers 3 users and 25 assets - register room that fits a complete single-site system’s devices.
FAQ
How do I keep an inventory of security cameras? One record per camera, recorder and switch: asset ID, location, model, serial, install date, warranty end, plus the recorder and channel each camera feeds. Then verify it on a schedule.
What should a CCTV asset register include? The full recording chain - cameras, NVRs, PoE switches, backup power - each with location, serial, warranty and a last-verified date.
How often should security cameras be checked? Monthly or quarterly walk-throughs for most sites: live view, clean lens, clear sightline, and a recorder retaining footage for the intended period.
Should NVRs and PoE switches be in the camera register? Yes - a dead switch or full recorder loses footage exactly like a dead camera, and the cabinet gets even less attention than the walls.
How do I label cameras that are mounted out of reach? Label the camera before installation, then put the working label at the patch panel or recorder port it feeds. Identification happens in the cabinet, not on the ladder.
The takeaway
Camera systems fail quietly, so the register’s job is to make silence visible. Inventory the whole chain down to the switch in the cabinet, keep the channel map current, label the reachable end of every run, and put a date on the last time each device was confirmed working. The goal is simple: the next time someone asks for footage, the answer is a clip - not an apology.