The one almost anybody knows how to maintain. And that is worth money.

It is an addressable control panel of the traditional sort: loops with detectors and call points identified one by one, supervised sounder lines, batteries, modules to talk to the rest of the building and the option of linking several panels together when there is more than one plant room. It belongs to a large group that also has other fire brands in higher ranges, and that shows in access to spares and in the use of very widespread detection protocols.

Where it turns up: offices, industrial units, logistics centres, hotels, shopping centres, care homes, administrative buildings. It is the workhorse of the mid range, and it is in a great many installations in this country. There is a fair chance that anyone buying a building put up in recent years will find it already on the wall.

That has a consequence that appears in no brochure and that is the best argument it has: there are a lot of technicians who know it. If tomorrow the client wants to change maintenance company, ask for a second quotation or a second opinion on a fault, they will find somebody. With uncommon electronics that does not always happen, and the client with no alternative ends up paying whatever they are told.

A fire system is a fifteen or twenty year contract, not a purchase. It is installed once and inspected every year, with its file, its certificates and its spares. The question almost nobody asks in the purchasing meeting is what happens if the relationship with whoever installed it comes to an end: over price, over service, or because the company leaves the business.

With a very widespread panel the answer is «you find another company». With a very rare one, or with one configured in a way only its author understands, the answer is «either you put up with it or you change the head end», and changing the head end of a working installation is a project with a shutdown, with every point checked again and with an invoice nobody had allowed for.

That is why we hand over the configuration file, the drawings of what is there and the programming passwords in the client's name, even when the client does not ask for them. It is not generosity: if the installation carries our name, it has to be possible to explain it and maintain it without us. The opposite is selling a system and keeping the key.

What you would expect, well resolved and with no surprises: loop cards added as the installation grows, a maximum number of devices and cable length per loop that it is best not to use up on handover day, power supply and batteries sized for the mains failure the project calls for, supervised relays and inputs for the ventilation, the doors and the lifts, and repeaters to take a screen to reception.

For detectors it takes the usual range: optical smoke, heat, dual technology for places with some dust or steam, and beam detectors for tall industrial units or large volumes. That variety is not a catalogue to pick from as you fancy: it is the toolbox so that each point carries what that point needs, which is the only decision that really decides whether the system is going to work.

The programming is done from a PC program and loaded into the panel, and what gets loaded is the cause and effect matrix: what triggers what. It is the most important document in the installation and the one most often inherited from another project without being reviewed. Going through it with the building's evacuation plan in front of you is what separates a system that was thought about from one that was copied.

What makes it reliable is maturity. It is electronics with many years of installations behind it, with the problems known and documented, and with technicians who have already run into what you are running into. In a system that only matters on one day, that weighs more than any new feature.

What makes it go off when it should not is the usual thing, and it is worth repeating because it is ninety per cent of the alerts raised for nothing: the detector badly chosen or badly placed. Optical smoke over a loading area, near a dock door, under a kitchen air outlet, in a changing room, in a workshop where welding is done. The panel has tools to compensate for it, and none of them fixes a detector that is in the wrong place.

It suits most ordinary buildings: one storey or several, a few hundred points, a need to know exactly which detector has gone off, and a wish not to depend on anyone in particular to maintain it. And it suits when it is already fitted, which is the most frequent case: often what is needed is not changing the panel, it is putting the installation you have in order.

It does not suit if what you need is to run dozens of panels spread across a campus with graphics on plan and redundancy in the network, nor if there is a technical room that calls for aspirating detection or a tunnel that calls for linear detection. Those are other products, they are in this same area, and forcing them into a mid-range panel does not turn out well and does not turn out cheap.

We are electronics people, and we say that up front. The detection, the panel, the call points, the sounders, the evacuation, the interfaces with the rest of the building, the tests and the maintenance: yes.

The hydraulic side, no. Sprinklers, dry risers, hose reels, pressure sets, tanks and their pipework are another trade with another licence, and anyone who tells you they do all of it had better explain to you under whose signature. Nor the mechanical side of gas suppression: the detection and the release, yes; the cylinders and their pipework, no.

The day to day is the client's: reading the panel, telling an alarm from a fault, silencing, resetting, isolating a zone because there is work going on and putting it back afterwards with a note. With training, an attentive maintenance manager sorts out most of the scares alone and stops ringing about things that do not deserve it.

The structural side, no: adding devices, changing the matrix, touching an interface or updating the program. Done halfway, the system stays lit and lying, which is the only fault you cannot see.

And here is the uncomfortable part, which is the same on all of them: the inspection schedule costs money every year and is not negotiable. What you can choose, and it is the gift a very widespread panel gives you, is who does it. If one year you are not convinced by whoever is doing it, you can change without changing the system. Very few technical decisions leave you that door open.