A security system is only truly useful if it keeps protecting the site when something goes wrong. Cameras, alarms and access control all sit on top of power, networks and communications, and those foundations are exactly what a power cut, a broadband fault or building works take away. Security resilience is the discipline of understanding those dependencies before they are tested for you.
Most conversations about school security are conversations about equipment. How many cameras, which access control system, whether the alarm is monitored. All reasonable questions. But they share a hidden assumption: that the electricity is on, the network is up, and the internet is working.
Security resilience asks the question underneath that assumption. If the power fails at 3am, or the broadband drops during the school day, or a contractor cuts through a duct during building works, what actually keeps working, and what quietly stops? A school can own a great deal of security technology and still have very little security resilience. The two are not the same thing.
Security resilience is the ability of a security system to keep providing useful protection when the conditions it relies on deteriorate. Not perfect protection, and not protection against every conceivable failure, but sensible, understood, planned-for continuity.
The distinction that matters is this:
A camera that records to the cloud is useful right up until the broadband fails, at which point it may be recording nothing. An electric gate is convenient until the power goes off and nobody knows how to release it. A monitored alarm is reassuring until its single signalling path goes down and the monitoring centre hears silence. None of these are exotic events. They are ordinary Tuesdays, and resilience is simply having thought about them in advance.
Schools are unusually exposed to this problem, for reasons that have nothing to do with threat levels and everything to do with how a school actually runs.
A school site is large, often spread across several buildings of different ages, with cabling and cabinets that have grown up over decades. It has long periods when almost nobody is on site: evenings, weekends and the school holidays, which is precisely when an outage is least likely to be noticed and most likely to matter. It runs a constant background of planned disruption too, from electrical maintenance and IT upgrades to refurbishments, temporary classrooms and building works, each of which can take a security system offline without anyone deciding that it should.
Resilience is not only about dramatic failures. It is just as much about the planned power-down for a rewire that nobody realised would take the CCTV recorder with it, or the network upgrade that quietly changed the path the alarm used to signal out. Good planning treats planned and unplanned disruption as the same discipline.
It helps to see a security system for what it really is: a stack of technology sitting on a stack of dependencies. The camera is the visible part. Underneath it sit power, cabling, network switches, recording equipment, internet connectivity, mobile signal, cloud services and a monitoring centre, and every one of those is a link that can break.
The useful habit is to look for single points of failure. If one power supply, one broadband line, one switch or one signalling path failing would take out something important, that is a single point of failure, and knowing where they are is most of the battle. You do not have to eliminate all of them. You do have to know they exist and decide, deliberately, which ones are worth doing something about.
To make this manageable, we find it helps schools to think in five layers. A security system is only as resilient as the weakest layer beneath it, so it is worth looking at each in turn rather than assuming the whole thing stands or falls together.
Everything starts here. Cameras, recorders, access control, gates, alarms and network equipment all need power, and they do not all fail gracefully when it disappears. Some cameras hold a battery-backed recorder for a while, some stop instantly. The question is not whether you have power, but what happens in the minutes and hours after you lose it, and which parts you have chosen to protect with an uninterruptible power supply (UPS) or, where it is justified, standby generation.
This layer is where cloud-dependent systems are quietly fragile. If your cameras record to the cloud and rely on broadband to do it, a connectivity fault can mean no recording at all, even though the cameras appear to be working. If your alarm or monitoring depends on a single internet or phone line, that line is a single point of failure. Resilient connectivity usually means recording locally as well as, or instead of, in the cloud, and having more than one way for a critical signal to get out.
The systems themselves, CCTV, access control, intruder alarms and the rest, each have their own failure behaviour, and it is rarely uniform. A door lock has to decide what to do when it loses power, and that decision is a life-safety matter, not a convenience one. A recorder has to keep footage safe through an outage. These behaviours should be designed in, understood and written down, not discovered during an incident.
Monitoring is only as good as the path between the site and whoever is watching. A monitored system that goes offline is worthless if nobody at the monitoring centre knows it has gone offline. Good monitoring detects loss of a system, not just alarms from it, and has a documented process for what happens next, including out of hours when the school is empty.
The final layer is the one technology cannot supply. When a system fails, someone has to notice, know what it means, and know what to do instead. Resilience that lives entirely in equipment is brittle. Resilience that also lives in people, in named responsibilities, current contact lists, manual procedures and a habit of testing, is the kind that holds up. This is where security resilience meets everyday preparedness.
A power cut is the clearest test of resilience, because it hits every layer at once. It is worth walking through calmly rather than assuming.
Cameras may stop immediately, or keep running if they and their recorder are on a UPS. Recording may continue or may halt, and the difference matters enormously after the event, when you go looking for footage that was never captured. Electric gates and doors will behave according to how they were configured, which for anything on an escape route is governed by fire safety, not preference. Intruder alarms should ride through a mains failure on their standby batteries for a period defined by their grade, then raise the alarm about the power loss itself. Access control may fall back to a safe default, but which default, and whether staff know it, varies from school to school.
None of this needs to be alarming. It needs to be known. Our companion piece, the Power-Off Checklist, goes through this layer in more depth.
Broadband failure is more common than power failure and often less noticed, because the lights stay on. The site looks normal. The question is what has silently changed.
If cameras record to a local recorder, they should keep recording through an internet outage, and you simply lose remote viewing until the line returns. If they record only to the cloud, you may be losing footage for the entire duration of the outage. Monitoring that depends on a single internet path may go blind, and whether anyone finds out depends on whether the monitoring centre actively checks that the path is alive. Systems that fall back to a mobile connection are more resilient, but only where there is reliable mobile coverage, which not every school has.
This is a planning aid, not a statement of legal requirements or a specification. Use it to understand your own dependencies and to have a better conversation with your IT team and your security provider. Treat every unticked box as a question worth asking, not a failing.
1. Power
2. Connectivity
3. CCTV
4. Access Control
5. Intruder Alarms
6. Monitoring
7. Network Infrastructure
8. People and Procedures
9. Testing
A resilience measure you have never tested is a hope, not a plan. The point of testing is not to prove everything works. It is to find the thing that does not, on a day you have chosen, rather than during a real incident.
Sensible tests include a planned power-down to see what really happens to CCTV, recording and access control; an end-to-end check that an alarm signal actually reaches the monitoring centre and gets the expected response; retrieving footage after an outage to confirm it was captured; and checking that backup power holds for as long as it is supposed to. The important discipline is that testing is planned, done safely, and carried out by appropriately qualified people. Powering down a live school site is not something to improvise.
A good test often produces an uncomfortable but valuable result: the footage you assumed was there is not, or the signalling path you thought was backed up is not. That is the test doing its job.
You do not need to be a systems engineer to have a productive conversation about resilience. A few plain questions reveal most of what matters:
A provider who can answer these clearly is thinking about your whole system. A provider who only wants to talk about cameras is selling you part of one.
Resilience is far cheaper designed in than bolted on. Whenever a school is planning a refurbishment, a new build, a network upgrade or a security refresh, that is the moment to ask the resilience questions, because the cost of running a second cable, protecting a cabinet or specifying local recording is trivial at design stage and awkward afterwards.
It is also the moment to think about systems together rather than separately. CCTV, access control, alarms, monitoring and the network are not really separate purchases. They share power, cabling and connectivity, and they succeed or fail together. Considering them as one system, with the dependencies mapped, is what turns a collection of products into something resilient.
Resilient security is not about predicting every possible failure or spending heavily to guard against events that may never happen. It is quieter and more useful than that. It is about understanding what your security depends on, finding the single points of failure, and making sensible, proportionate preparations for the ordinary disruptions that every school eventually meets.
The schools that cope well with an outage are rarely the ones with the most equipment. They are the ones who understood their dependencies before those dependencies were tested for them.
If your school or trust is reviewing its security, or planning works that will touch it, Fyrfly can help you look at the whole system, power, connectivity, systems, monitoring and people, and find the points where resilience is worth building in.
A free Fyrfly site survey maps how your power, connectivity, systems, monitoring and people work together, and shows you where the single points of failure are. Practical, honest, and no fear-selling.
Request a Free Survey →