Your alarm panel does not watch your doors. It watches wires.
That sentence is the whole subject. Every contact, motion detector and glassbreak in the building reports through a pair of conductors running back to a panel, and everything the panel knows about your building it infers from the electrical condition of those conductors. The detector is a switch. The wire is the message. The panel is a device that reads wires and decides what they mean.
Which raises a question almost nobody asks when buying a system: what, exactly, can the panel tell apart?
The unsupervised circuit and its single fact
Take the simplest possible arrangement — a door contact wired to a panel input, nothing else on the run.
The contact is closed when the door is shut and open when the door is opened. The panel is watching for continuity: is the circuit complete, or is it broken? Complete means the door is shut. Broken means the door is open. That is the entire vocabulary of the arrangement. One bit of information, two possible values.
Now consider everything else that produces a broken circuit.
A staple driven slightly too hard during a renovation, pinching a conductor until it parts months later. A run passing over a sharp edge in a ceiling space, chafing through with building movement. A splice in a damp location corroding open across a winter. A conductor pulled apart when a contractor moved a door frame. A detector that lost power. A rodent.
Every one of those produces the same reading as an opened door — and if the system is disarmed at the time, which it is for most of the week in most buildings, that reading produces nothing at all. The event is not an alarm. It is not a trouble. It is not anything. The zone simply stops being able to report, and no part of the system is constructed to notice.
That is the failure mode worth understanding, because of the direction it fails in. The zone does not fail loudly. It fails into a state that is indistinguishable from a quiet building, on a system whose keypad continues to say Ready and whose owner continues to arm it every night with complete confidence. A detector that has been electrically dead since March will not be discovered in March. It will be discovered during a service call in November, or by an incident, or never.
What supervision changes: measurement instead of continuity
A supervised loop replaces the yes-or-no question with a measurement.
The change is small and the consequence is not. Instead of asking "is this circuit complete," the panel measures the electrical resistance of the loop and compares it against what it expects to see. A resistor of a known value — the end-of-line resistor — is placed at the far end of the circuit, at the last device, so that a healthy loop presents a specific, expected value back at the panel rather than merely being unbroken.
Now the panel has more than two readings to work with, and each one means something different:
Normal. The measured value matches what a healthy, secure loop should present. Devices are connected, the wiring is intact, everything is where it should be.
Alarm. The value has changed in the way a detector operating is supposed to change it. The door opened, the motion detector tripped, the contact did what contacts do.
Open circuit. The panel sees no complete path at all — a reading that a working loop cannot produce, whatever the doors are doing. A conductor is severed, a terminal has come loose, a device has been removed, or a splice has corroded through.
Short circuit. The path is complete but the resistance is gone — the loop is reading as though the far end of it does not exist. Two conductors are touching where they should not: crushed cable, a screw through a run, water in a junction, insulation degraded to the point of contact.
The last two are the ones you are paying for. They are not alarms. They are the system reporting on itself — a trouble condition, meaning the panel has concluded that this zone is no longer able to do its job and is telling you so, tonight, rather than leaving you to find out later.
That is the entire argument for supervision, and it is worth stating without ornament: an unsupervised zone reports on your building; a supervised zone reports on your building and on itself.
Why a system that reports its own faults is a different class of system
The instinct is to file supervision under tamper detection — the idea that it is there to catch somebody deliberately cutting a wire. It does contribute to that, and we are deliberately not going to describe how any circuit is defeated, which is a subject that belongs between a system designer and the standards their installation is built to, not on a public page.
But framing supervision as an anti-tamper feature undersells it, because deliberate attacks on alarm wiring are rare and wiring faults are not. The everyday value of supervision is reliability engineering. It is the difference between a system with a health signal and a system without one.
Think about what you actually rely on your alarm for. Not the siren — the siren is the last few seconds of a long chain. What you are relying on is the coverage: the assumption that every opening and every space you decided to protect is currently being watched. Every one of those detectors is a small piece of hardware at the end of a long conductor in a building that gets renovated, cleaned, flooded, drilled into and rearranged. Coverage decays. It decays quietly. The only question that matters is whether the decay is visible.
Supervision is what makes it visible. It converts a silent failure into a signal that appears on a keypad, in a monitoring centre, and in a log — while the building is otherwise fine, at a moment when fixing it is a service call rather than an incident report.
This is the same property that separates modern access-control wiring from the legacy interface most BC buildings still run between reader and controller. The industry's own replacement for that interface is named for the feature: the S in OSDP stands for supervised, and its practical benefit is that the controller notices when a reader stops answering instead of treating silence as normal. We covered that in the wire between your card reader and the door. The mechanism is different; the principle is identical. A system that cannot detect its own faults will not report them, and a fault that is not reported is a fault you own without knowing it.
The states are only half of it — the other half is where the signal goes
Here is where supervision gets quietly undone in real buildings.
A trouble condition is not an alarm. It does not sound a siren. It typically produces a small indication on a keypad and, if the system is monitored and configured to transmit it, a low-priority signal to the monitoring centre. Which means supervision can be fully functional and still deliver you nothing, if the trouble signal terminates somewhere nobody looks.
The failure patterns are consistent and they are all administrative rather than technical: the panel indicates a trouble that staff have learned to ignore because it has been showing for months. The monitoring contract covers alarm signals but the handling of trouble signals was never discussed. The notification goes to an email address belonging to a property manager who left the company two years ago. The system is monitored, the trouble is transmitted, it is logged, and no human ever converts it into a work order.
A supervised system whose troubles land in an unwatched log is an unsupervised system with better documentation. Supervision is a chain — detect, transmit, deliver, act — and the technical part of that chain is the part most likely to be working.
The related discipline is testing, because supervision tells you a zone is electrically healthy, not that a detector is aimed correctly or that a contact still lines up with a door that has sagged on its hinges. Systems degrade in both directions and only one of them shows up as a trouble. If you have not read it, what a maintenance obligation actually looks like in BC covers the documentation side of that, and the same reasoning applies here — the difference between a system and a maintained system is a record.
What to ask, and how to know
This is the owner-facing part, and it is short because the questions are simple. Ask them of whoever maintains your system, in writing, and keep the answers in the building file.
1. "Are all of my zones supervised, or only some of them?" Mixed installations are extremely common — the original panel's zones one way, an addition wired another way, a retrofitted device on a spare input that nobody set up properly. The answer is very often "most of them," and the value is entirely in knowing which ones are not.
2. "Show me the zone list, and mark which zones are supervised." A list, per zone, per door. Not a reassurance. If the person maintaining your system cannot produce this from the panel programming, that is itself the finding.
3. "If a detector stopped reporting tonight, what would happen, and who would find out?" This is the question that does the most work, because it tests the whole chain rather than the wiring. You are looking for a specific answer: it appears as a trouble on the keypad, it transmits to the monitoring centre, the centre contacts this named person, and that person opens a service call. Any vagueness in that sequence is where your supervision actually ends.
4. "When was the last time a trouble condition was raised and closed on this system?" A system that has never reported a fault in years is either in unusually good condition or is not capable of telling you. Those look identical from the keypad. The service history separates them.
5. "What is the current trouble state right now?" Ask during the walkthrough. Standing troubles that everyone has normalized are the most common finding on an older installation, and they are frequently the reason a subsequent real fault went unnoticed — the indicator was already on.
The related conversation, if your site generates them, is nuisance signals — because a system reporting faults it does not have trains people to ignore the ones it does. That is a separate subject and we have written about what false alarms cost in BC elsewhere.
The short version
Your panel infers everything it knows about your building from the electrical condition of a pair of wires. If it only checks whether those wires are unbroken, it knows one thing, and every silent wiring failure in the building reads to it as a perfectly normal night.
A supervised loop, with a resistor of a known value at the far end of the run, lets the panel measure instead of merely check — and the difference between "normal," "alarm," "open" and "short" is the difference between a system that can report its own decay and one that cannot. Supervision does not make your alarm more sensitive. It makes it honest about whether it is still working.
Then ask the second half of the question, which is where most buildings actually lose the benefit: when this system reports a fault, whose desk does it land on?
