Building automation is not just for towers, and the cheapest win is a schedule

The short version

• Scheduling is the cheapest change and usually the largest saving.
• You cannot manage what you cannot see, so metering by zone comes before optimisation.
• Occupancy sensing works well in some spaces and badly in others, and the difference is predictable.
• Anything with a heating element is worth more attention than the lighting.
• Controls added to badly maintained equipment automate a problem rather than fixing it.

Start with the schedule, because it is free

Walk your building at 9pm on a Sunday. Note what is on.

In most small commercial buildings the answer includes lights in areas nobody enters after 5pm, HVAC maintaining a comfortable temperature for an empty room, and a rooftop unit cycling because its thermostat has been on the same setting since it was installed.

Almost every thermostat and lighting panel made in the last twenty years supports scheduling. Most are not scheduled, because setting it up is a job nobody was assigned and because the person who installed it left it on a default.

Before spending anything, someone should sit down with every thermostat and panel in the building and set occupied and unoccupied periods that match when people are actually there. That includes the setback temperature, which is the point of the exercise. A building held at 21 degrees overnight costs meaningfully more than one allowed to drift to 16 and brought back up before people arrive.

Two things to get right, because getting them wrong produces complaints that kill the whole project.

Recovery time. The system needs to start early enough that the space is comfortable when the first person arrives, not at the moment they arrive. Cold mornings need a longer lead.

Freeze protection. Any setback has a floor. Pipes in an unheated area, sprinkler lines, anything with water in it. This is where a schedule set carelessly causes real damage, and it is worth being explicit about rather than assuming the equipment handles it.

Metering before optimising

The reason building automation projects disappoint is that people install controls without knowing where the energy goes, then cannot tell whether anything improved.

Sub-metering by zone or by system gives you a baseline. It also routinely finds something surprising. A compressor running continuously because a valve is stuck. A heater in a small room that draws more than the entire lighting circuit. An outbuilding on the same meter that everyone forgot about.

Anything with a heating element deserves particular attention: unit heaters, hot water tanks, electric baseboards in a back office, heat tracing on a pipe. Resistive heat is the most expensive way to make warmth, and it is often on a switch someone flipped in November and never revisited.

You do not need permanent metering everywhere to start. A clamp meter and an afternoon tells you a great deal about a small building.

Occupancy sensing, and where it fails

Sensors that turn lights off in empty rooms are genuinely effective in some spaces and irritating in others.

Works well: washrooms, storerooms, stairwells, meeting rooms, corridors. Places people pass through or occupy briefly, where a light left on is pure waste.

Works badly: any space where somebody sits still. An office where someone works quietly at a desk will have the lights turn off on them, which is the single fastest way to make staff hate a new system. The technology matters here, since some sensors detect motion only and others detect presence more reliably, but the safer answer in a private office is a longer timeout or nothing at all.

Also poor: warehouses with high ceilings and tall racking, where a sensor’s view is blocked and someone working in an aisle is invisible to it. High-bay spaces need sensors specified for the mounting height and the aisle layout, not the same units used in an office.

There is a related trap in cold storage and in any space with equipment that needs air movement. Turning off ventilation to save energy in a room that needs it for another reason is not a saving.

Where this touches access control

If you already have access control on your doors, the building knows when the first person arrived and when the last one left. That is useful information for a schedule, and in some systems the two can be linked so the building responds to actual occupancy rather than an assumed pattern.

That integration is worth asking about, because it is the difference between a schedule that fits a fixed nine-to-five and a building that adapts when a crew starts at six.

Remote access, and its security question

Being able to see and adjust the building from a phone is the feature people want most, and it is genuinely useful. It stops the drive-in on a Sunday to reset a thermostat.

It also puts a controller on your network reachable from outside, and building controls are not known for strong security. Many run old firmware that will never be updated.

Two rules worth applying. Controls go on their own network segment, separated from the network your computers and files live on. And remote access goes through a controlled path with proper authentication rather than by exposing the controller’s own web page to the internet. A building controller found by an internet scan is a real category of incident, not a theoretical one.

What this means for your building

A warehouse in Courtice with unit heaters and high-bay lighting has the largest available saving, and most of it is in scheduling and in getting the high-bay sensors specified correctly for the racking layout.

An office in Ajax across two floors usually finds its win in zone control, because one thermostat serving a floor with a sunny side and a shaded side keeps half the staff uncomfortable and the system working hard.

A clinic in Whitby has constraints rather than opportunities in some rooms, since temperature and ventilation requirements limit what can be set back. The savings are in the areas without those requirements: waiting room, offices, corridors.

An older building anywhere should get its equipment serviced before its controls upgraded. Automating a rooftop unit with a failing damper produces a well-scheduled failure.

What to do about it

1. Walk the building on a Sunday evening and write down what is running. This is your business case, and it costs an hour.
2. Set schedules on every thermostat and lighting panel you already own, with sensible recovery time and a freeze protection floor.
3. Find the resistive heat. Anything with an element. Check whether it needs to be on.
4. Meter or clamp-test the major loads before buying controls, so you know where to aim and can prove the result.
5. Service the mechanical equipment before adding controls to it.
6. If controls go on the network, segment them and keep remote access behind proper authentication.

Lighting and HVAC integration, energy monitoring by zone and remote management is one of the things we do, and the network and security side comes with it rather than being someone else’s problem.