The Plant Is Usually Fine. The Sequence Is Not.
Most underperforming halls we survey are not short of cooling capacity. They are running adequate equipment on a poor control strategy — CRAC units humidifying while the unit beside them dehumidifies, fans holding a pressure nobody specified, and setpoints inherited from a commissioning sheet written before the current IT load existed. None of that shows up as a fault. It shows up as energy cost, and as capacity that appears unavailable when it is not.
The UAE climate narrows the margin further. High ambient temperatures reduce economiser hours and push heat rejection onto chiller plant or a district cooling connection, so the sequencing between facility side and white space has to respond to live conditions rather than run fixed. Where liquid-cooled AI rows are being added alongside legacy air-cooled halls, that coordination becomes the whole job — see liquid cooling controls for the high-density side.
Our position is specific: Intellicon is the controls and integration layer, not a DCIM vendor or an equipment supplier. We integrate what your mechanical, electrical and IT contractors install, and we own the points, graphics, trends, alarms and interlocks around it.
What We Monitor and Control
CRAC and CRAH sequencing
Duty and standby rotation, staged capacity against live load, and coordinated humidity control so units stop working against each other.
Containment differential pressure
Continuous pressure measurement between contained aisle and room, with fan speed modulated to pressure and rack inlet rather than a fixed return-air setpoint.
Rack-level environmental data
Inlet temperature and humidity trended per rack or per row, so hot spots are visible individually instead of averaged into a single hall reading.
PUE and energy reporting
Continuous PUE from real intake and IT-side metering, trended against ambient and load — or a metering plan first, where the meters do not yet exist.
Unified alarm hierarchy
One severity scheme and one escalation path across cooling, power and leak detection, so nothing depends on which console an operator happens to be watching.
Power chain interface
UPS, generator and PDU status brought onto the same head-end, with power quality and harmonics monitoring delivered alongside our sister company King Link Electrical.
Two Systems, One Set of Numbers
The BMS owns the plant. Cooling units, chilled water, pumps and valves, containment pressure, environmental conditions, leak zones and the interlocks that act automatically when something fails. It runs continuously, and it has to be right at three in the morning with nobody watching.
DCIM owns the IT estate. Rack and asset inventory, port and circuit capacity, power draw per cabinet, and the change process around installing or decommissioning hardware. It is a planning tool more than a control system.
The overlap is power and thermal data — and that is where sites end up with two platforms quoting different figures to different people. The integration work is deciding which system is authoritative for each value, then making the other consume it rather than measure it independently.
For mixed or ageing estates the practical starting point is a survey of the existing head-end and points list. Our BMS retrofit and integration and BACnet and Modbus integration pages cover how we approach systems that were never designed to talk to each other.
Frequently Asked Questions
What is the difference between a BMS and DCIM, and does a data centre need both?
They answer different questions. The BMS owns the physical plant — cooling units, chilled water, pumps, valves, containment pressure, ventilation and the environmental conditions in the white space. DCIM owns the IT estate — asset and rack inventory, port and circuit capacity, power draw per rack, and change management. The overlap is power and thermal data, which is exactly where most sites end up with two systems reporting different numbers. In practice a facility needs the BMS to run the plant reliably and DCIM to plan capacity, with a defined integration between them so that one set of figures is authoritative. We build the BMS layer and the integration; we do not replace DCIM.
What does the BMS actually control in a data centre white space?
Cooling unit sequencing and staging, so that CRAC and CRAH units share load rather than fighting each other with opposing humidification and dehumidification cycles — a common and expensive fault we find on existing sites. Beyond that: containment differential pressure, chilled water supply temperature and flow, valve control, rack inlet temperature and humidity trending, leak detection zones, and the interlocks that decide what happens automatically when a unit fails. The control strategy matters more than the hardware; most underperforming halls we survey have adequate equipment running a poor sequence.
Can the BMS calculate and report PUE automatically?
Yes, provided the metering exists to support it honestly. PUE is total facility energy divided by IT equipment energy, so it requires metering at the utility intake and at the UPS output or PDU level. Where those meters are present we bring them onto the head-end and produce continuous rather than annual PUE, trended against ambient temperature and IT load so the number is interpretable. Where the metering is not present, the first deliverable is a metering plan — a PUE derived from an estimated IT load is not a measurement, and we will say so rather than produce a flattering figure.
Can you integrate with our existing DCIM or monitoring platform?
In most cases yes, subject to a protocol assessment during site survey. We work vendor-neutral across Siemens, Honeywell, Schneider and open-protocol platforms, and the practical route is usually BACnet/IP or Modbus TCP, with SNMP where the IT side prefers it. The objective is a single alarm hierarchy rather than two consoles showing overlapping alerts at different severities, because split alarm ownership is what causes real events to be missed during a shift handover.
What does hot and cold aisle containment need from the controls side?
Differential pressure control between the contained aisle and the room, measured continuously rather than commissioned once and assumed stable. Containment changes the airflow relationship, so fan speed has to modulate against pressure and rack inlet temperature instead of against a fixed return-air setpoint. Without that, containment can either starve the racks or waste significant fan energy holding a pressure that is not needed. This is one of the most common retrofit opportunities we find in existing UAE facilities.
Does Intellicon supply DCIM software, cooling units or IT hardware?
No. Intellicon is the controls, monitoring and integration layer. We do not manufacture or resell CRAC units, chillers, DCIM platforms, servers or racks. We integrate the plant and systems your mechanical, electrical and IT contractors install into a coherent BMS — points, graphics, trends, alarms, interlocks and reporting — and we own that layer. Scope is always confirmed by site survey and design review before anything is quoted.
Reviewing a Data Centre Controls Scope?
We will review your existing head-end, points list and metering, then scope the integration honestly.
Related solutions and guides
- Liquid Cooling Controls for Data CentresCDU monitoring and leak detection for high-density racks.
- Custom Transformers for Data CentresIsolation, cast resin and K-rated transformers.
- Custom Voltage Stabilisers for Data CentresServo, static and three-phase stabilisation.
- Guide: BMS for Data Centre Cooling ControlPrecision temperature and power monitoring in UAE data halls.
Want to know what is reachable in your building? Request a data-centre controls proposal