Plain language
What a building management system controls when nobody is watching
We do not have an office in Saudi Arabia
We serve organisations there from our Riyadh office and our engineering hubs in Lahore and Karachi, travelling for the parts of an engagement that need to happen in person.
Strip the acronyms away and a BMS is the building’s reflexes. It starts the chillers before the morning heat arrives and stages them as the load climbs; it holds each floor’s temperature by modulating dampers and valves; it schedules lighting to occupancy and daylight; it watches pumps, fans and tank levels and raises alarms when they drift; and in many estates it coordinates with access control and fire systems so that doors, air and alarms behave coherently in an emergency. All of this runs from controllers wired to thousands of field devices — sensors, actuators, meters — with a head-end workstation, usually in the plant room, as its face.
That closed competence is precisely why the BMS was, for decades, nobody’s problem but the controls contractor’s. It is also why it is now everybody’s interest: the system that runs the largest energy loads in the building, knows which spaces are conditioned for whom, and sees every mechanical fault first, holds data the estate’s operators, owners and sustainability reporting all want. The engineering question of this decade is not whether to connect it, but how to do so without inheriting the risks the isolation was hiding.
The integration
From a valve in the plant room to a figure the enterprise can use
The stages a building signal passes through on its way to enterprise systems. Each boundary exists for a reason, and the design work is deciding what crosses each one, in which direction.
A flow from field devices and controllers, through the BMS head-end and a segregated integration layer, to enterprise systems — with a branch showing local control continuing if the link fails.
- 01
Field devices and controllers
Sensors and actuators wired to controllers that run the actual control loops — the layer that must keep working whatever happens above it.
- 02
The BMS head-end
Where the building’s points, schedules and alarms are supervised — historically the end of the line, now the source of record.
- 03
The integration layer
A read-first gateway that speaks the controls protocols on one side and enterprise interfaces on the other, across a segregated network boundary.
- 04
Enterprise systems
The FM platform raising work orders from alarms, the energy reporting the owners read, the analytics that challenge schedules against occupancy.
Local control survives the link
The controllers keep running the building if the integration layer, the corporate network or the cloud disappears. Any design in which a WAN outage can affect a control loop has put the dependency in the wrong direction.
The retrofit reality
Wiring intelligence into a building that was finished years ago
New builds arrive with integration in the specification. Most of the estate that exists did not, and these are the conditions a retrofit actually meets on site.
- 01
Drawings that describe a different building
Years of maintenance, tenant fit-outs and undocumented changes separate the as-built drawings from the installation. The retrofit begins with a survey of what is actually wired, because integrating a documented fiction wastes everyone’s season.
- 02
A head-end only one firm can talk to
Older installations often expose their points only through the original vendor’s tools and the contractor who holds the passwords. Part of the work is commercial — establishing access and export rights — before any of it is technical.
- 03
Tenants in residence while you work
An occupied tower allows no commissioning downtime: changes happen system by system, at night or riser by riser, with the ability to fall back at every step. The schedule is set by occupancy, not by engineering preference.
- 04
A budget that must pay as it goes
Few owners fund a whole-estate retrofit up front. The programme is sequenced so early phases produce usable results — the energy view, the alarm-to-work-order path — that justify the next phase from evidence rather than faith.
The data dividend
Years of BMS trend logs are an asset most estates delete
A BMS records constantly — temperatures, runtimes, alarms, setpoint changes — and in most buildings that history is overwritten on a cycle or trapped on the plant-room workstation until a disk fills. Kept, contextualised and joined to the estate’s other records, it becomes operational memory: the chiller whose approach temperature has drifted for a month before it fails, the floor whose after-hours consumption says its schedule no longer matches its tenants, the alarm pattern that distinguishes a failing sensor from a failing pump. None of this requires new instruments; it requires treating what the building already writes down as data the operation owns, stored where analysis can reach it, under a retention the owner chooses rather than a controller’s buffer size.
The same connection that carries the dividend carries the risk, which is why the two belong in one design. A building network reachable from corporate IT is reachable, in principle, by whatever compromises corporate IT — and controls installations were engineered for reliability, not for hostile traffic. The connection is therefore built as a guarded boundary: segregated networks, a one-way preference for data flowing out, remote access for contractors that is granted per session and logged, and the enterprise side treating the building feed as it would any external source. Estates whose buildings include critical facilities should treat their controls with the same discipline as any operational technology; our data-centre OT security work describes that discipline at its most demanding.
Our role
What we engineer between the BMS and IT, and what the controls contractor keeps
- 01
The survey and the access
We establish what each building’s installation can actually expose — protocols, points, export rights — and put the commercial access questions to the incumbent contractor early, because they decide the schedule more than the technology does.
- 02
The guarded connection
We design and build the integration layer and its network boundary with your IT and security teams — segregation, monitored flows, per-session contractor access — so the estate gains the data without gaining an attack path.
- 03
The consumers that justify it
We connect the flow to the systems that make it worth having — the FM platform’s work orders, the energy reporting, the analytics — because an integration whose data nobody consumes is a risk with no return.
The control loops, the safeties and the mechanical plant stay with the BMS and the contractors who maintain it — we do not touch what keeps the building safe; we build the layer that lets the rest of the enterprise learn from it.
Common questions
What buyers ask
Do we have to replace the BMS before connecting it to enterprise systems?+
Rarely, and a replacement justified only by integration deserves suspicion. Most installations of the last two decades speak standard controls protocols or can expose their points through a gateway, and the integration layer is designed to meet the system where it is. The genuine exceptions are very old or fully proprietary head-ends with no readable interface and no vendor path to one — and even there, the alternative to replacement is often a partial route: metering and selected points brought out independently while the controls keep running untouched. The survey answers this per building before anything is promised, which is exactly why it comes first.
Who should own responsibility for the security of building control networks?+
Named people on both sides of the boundary, under one agreed design — the failure mode is not the wrong owner but the gap between two. The controls contractor typically owns the devices and their maintenance access; corporate security owns the network policy and monitoring; and the estate’s operator owns the decisions about what connects at all. What we bring is the design that gives each party a defined surface: the contractor’s remote access granted per session and logged instead of a standing modem nobody remembers, the security team seeing the building networks in their monitoring rather than treating them as someone else’s plumbing, and changes to the boundary going through both. An estate that cannot name who owns this today has answered the question by default, and not well.
What is BMS trend data actually good for outside the plant room?+
Three uses pay reliably. Maintenance: equipment drifting from its normal behaviour — longer runtimes, widening temperature approaches, rising alarm frequency — shows in the trends weeks before failure, early enough for a planned intervention instead of an emergency one. Energy: consumption joined to schedules and occupancy exposes the floors conditioned for nobody and the plant staged out of habit, which is where an estate’s cheapest savings sit. Accountability: owner reports, sustainability disclosures and service-level discussions read from recorded history rather than recollection. All three depend on the same unglamorous foundation — the data kept beyond the controller’s buffer, named consistently, and stored where the operation’s tools can reach it.
In this cluster
Next step
Walk it through your estate
A survey of one building’s controls — what the BMS can expose, what state its points are in, and the safest path for its data to reach the enterprise.

