Smart Buildings: How IoT Cuts Costs and Emissions

Here is an uncomfortable number: the average commercial building wastes roughly a third of the energy it consumes. Lights burning in empty rooms. Heating running full tilt three hours after the last employee left. Air conditioning and heating fighting each other because two thermostats disagree on what the building should be doing. Nobody is ignoring this because they do not care — they are ignoring it because fixing it used to mean expensive, proprietary systems and year-long installation projects. That has changed.

What a "Smart Building" Actually Is

Strip away the marketing and a smart building is just this: sensors that measure what is happening, software that makes decisions based on those measurements, and devices (heating, lighting, ventilation) that act on those decisions — without a human clicking anything.

It is not a product you buy. It is a control layer added on top of the building systems you already have. Most of the physical infrastructure — heat pumps, ventilation units, lamps, boilers — can stay. What changes is how they get controlled.

Where the Waste Actually Comes From

Three things account for most of the energy waste in commercial buildings. None of them are exotic.

Heating and cooling space nobody is using

Meeting rooms booked for 9 AM but used at 10:30. Floors where half the team is remote on a Friday. Buildings that pre-heat on the same fixed schedule whether it is -15°C or +5°C outside. A simple occupancy sensor combined with a weather forecast solves most of it.

Lighting on dumb schedules

Lights that switch on at 7 AM and off at 7 PM regardless of who is in the building or how bright it already is outside. Daylight sensors and motion detection typically cut lighting electricity by 30–50% with no loss of comfort. Nobody notices except the accountant.

Running expensive equipment when electricity is expensive

In Latvia and most of Europe, electricity prices now swing by a factor of five or more within a single day. A dumb building pays whatever the rate happens to be. A smart building pre-heats hot water at 3 AM when it is cheap, shifts heavy loads away from peak hours, and spends up to 20% less on the same kWh of energy.

What We Have Actually Built

This is not theory. At BITBOX.lv we have built these systems in production:

Adaptive street lighting (with Idea Lights)

Motion-sensitive streetlights that run dim when the street is empty and brighten as a person or vehicle approaches. Municipalities using the system report up to 90% lower lighting electricity costs compared to running at full brightness all night. The streets are still safe, still lit, still functional — they just stop wasting power on empty roads at 3 AM.

Office energy management (with TET and Riga Technical University)

A platform that watches electricity spot prices, solar generation, and actual occupancy, then schedules heating, cooling and EV charging around the cheapest and cleanest windows of the day. For a mid-sized office, the savings typically cover the entire system cost inside two years.

The Integration Problem — and the Open Answer

Most buildings are a mess of systems from different vendors. HVAC from one company, lighting controls from another, security from a third — none of them talk to each other, and each vendor wants to sell you a proprietary “platform” to tie them together. Which locks you in forever.

We prefer to avoid that trap. Open platforms like Home Assistant, combined with custom integrations for the specific hardware a building already has, give you a single control layer without chaining you to one vendor. Five years from now when a piece of hardware needs replacing, you swap the hardware — not the entire system.

Where AI Actually Helps

A lot of the decisions a building needs to make are boring — turn off the lights when the room is empty, drop heating by 2°C at night. Those are not AI problems. They are if-this-then-that rules and they work fine.

AI matters when the decisions get harder. Predicting how quickly a room will warm up given today's weather, yesterday's usage pattern, and tomorrow's forecast. Deciding when to run a heat pump given variable electricity prices over the next 24 hours. Learning that the top floor tends to overheat on sunny afternoons and pre-cooling it two hours in advance. That is where a model genuinely beats static rules.

What the Numbers Usually Look Like

For a typical office building, a well-designed retrofit delivers 20–30% lower total energy bills, 30–40% less electricity on lighting specifically, a 15–25% drop in carbon emissions, and — usually the most underrated benefit — fewer “too hot / too cold” complaints from the people actually working there.

Payback periods sit in the 2–4 year range, and the hardware keeps running long after that.

What Implementation Actually Looks Like

For a mid-sized building the work breaks down like this: a walkthrough to map existing systems and find the biggest inefficiencies, installation of sensors (most are wireless — no cable runs through finished walls), setting up the control platform that ties everything together, tuning rules and models for how your specific building actually behaves, and ongoing monitoring to catch drift.

Total elapsed time is usually 4–10 weeks, depending on building size. No need to shut anything down while the work happens.

Why This Matters Beyond the Bill

Latvia's national energy targets, EU directives on building efficiency, and increasing ESG pressure from tenants, banks and investors all point the same direction. Buildings that do not modernise will face higher operating costs, worse rental yields, and eventually regulatory penalties. Buildings that do modernise are cheaper to run, easier to lease, and worth more when sold.

Getting Started

If you manage or own a commercial building and want to know what is realistic — what can be done, what it costs, what returns to expect — we will do an assessment and give you a straight answer. No proprietary platform pitch, no decade-long commitments. Just the real picture for your specific building.

Get in touch and let's take a look.