HVAC Optimization with Occupancy Sensors Explained

Most commercial HVAC systems still run on set schedules. They heat, cool, and ventilate entire floors based on when people are expected to be there, not when they actually are. 

‍

In a hybrid work environment, daily occupancy can swing from 30% to 80% with little warning. That gap between schedule and reality translates directly into wasted energy, unnecessary wear on equipment, and spaces that are either too warm or too cold.

‍

Occupancy sensors have become one of the most effective solutions to this problem. By feeding real-time presence and headcount data into building management systems, they allow HVAC to respond to actual demand rather than assumptions. 

‍

That means less energy spent conditioning empty space, better comfort where people actually work, and operational savings that compound across a portfolio.

‍

Why Schedule-Based HVAC Wastes Energy

Heating, cooling, and ventilation make up the largest share of energy use in commercial buildings. Space heating alone is the single biggest end use at about 32% of the total, with ventilation adding another 10%, according to U.S. Energy Information Administration data. So the schedule that controls those systems ties directly to the utilities bill.

‍

A building management system (BMS) typically runs HVAC on time-of-day schedules with fixed setpoints. The approach assumes occupancy stays consistent and predictable from one week to the next.

‍

Hybrid work breaks that assumption, because the number of employees on site now changes day to day. A floor with a total capacity of 200 might hold 60 employees and visitors on a Monday and 180 on a Wednesday, and a scheduled system treats both the same.

‍

That energy goes to the wrong places:

• Empty zones get the same ventilation rate as full ones. Fans and chillers keep working to condition air no one is breathing.
• The system ramps to full output at a set start time, whether the building is 20% or 90% full. A quarter-full Monday draws nearly the same energy as a packed Wednesday.
• After-hours and weekend activity in 24/7 global offices falls outside the schedule, so HVAC doesn't respond. People working late get stuck with stuffy air, or override the system and run a whole floor for a few people.

‍

The cost runs both ways. A half-empty floor keeps getting conditioned as if it's full, while a packed conference room can't get more fresh air until the next scheduled change.

‍

The BMS has no read on how many people are in a zone at any given moment. Giving it that data solves both problems.

‍

Many buildings have tried motion sensors, but the readout has a ceiling. Passive infrared (PIR) sensors detect movement without counting who's in a room. The BMS learns someone is present and still has to guess whether to ventilate for two people or 20.

‍

How Occupancy Sensors Optimize HVAC Systems

Occupancy sensors supply that read by reporting how many people are in each zone in real time. Instead of running on a clock, the BMS runs on what's happening in the building at that moment.

‍

The Basic Mechanism

An occupancy sensor reports headcount and zone status to the BMS every 30 to 120 seconds. The system compares that live count against its schedule and setpoints, then adjusts on the spot.

‍

When a zone is empty, the BMS widens the temperature deadband, the range it allows before heating or cooling kicks in. Stretching that band from 21 to 23°C out to 18 to 26°C keeps the compressor idle longer and cuts energy use. 

‍

Say your east wing clears out after lunch. The system lets that zone drift a few degrees instead of holding a tight 22°C for empty desks. When someone badges in at 3 PM, it pulls the temperature back before they notice. When the same zone fills, the system pre-conditions it before anyone feels the difference.

‍

Zone-Level Control

Whole-building control treats every floor as one unit, so a single busy area can pull the entire system to full output. Zone-level data changes that, letting the BMS run separate profiles at once. When one floor hosts an all-hands while three others are nearly empty, it conditions one floor instead of four.

‍

The same granularity tracks how a single floor shifts through the day. Workstations and open seating fill gradually through the morning, while meeting rooms spike and empty in 30-minute bursts. A room booked but unused still gets heated and cooled on schedule, and zone-level data lets the system stand down instead of conditioning it for a meeting that no one attends.

‍

Demand-Controlled Ventilation

Occupancy data drives ventilation on the same principle. The American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) sets minimum outdoor-air rates for indoor air quality in its Standard 62.1. Part of that minimum scales with how many people a space is built to hold.

‍

A room designed for 30 people draws fresh air for a full house, even when it's empty. But demand-controlled ventilation (DCV) ties the per-person share to the live count instead. When a zone runs light, the system eases off the intake and saves fan energy.

‍

When a meeting fills a room, DCV raises the supply in step, so air quality holds without waiting for carbon dioxide (CO2) levels to climb. The per-person share flexes with occupancy, though it can't drop below the baseline the standard sets for the space itself.

‍

Real-time adjustment, zone-level profiles, and DCV all draw on one input, a live count of who's in the building. With that count, the BMS stops heating and cooling the building it expects and starts heating and cooling the building that's there.

‍

What HVAC Optimization Looks Like in Practice

Across a portfolio of buildings, the gap between a fixed timetable and live occupancy data shows up directly on the energy budget.

‍

A global technology leader with more than 60,000 employees rolled out occupancy sensors across its high-traffic Bay Area office spaces. Its workplace and real estate teams had tried other solutions first. Some lacked accuracy in real conditions, and others raised privacy or cost concerns.

‍

The brief was tight. One sensing platform had to come online in under six weeks. It also had to cover four jobs at once across HVAC and cleaning efficiency, long-term space planning, high-traffic management, and round-the-clock operations.

‍

HVAC and Operational Outcomes

The deployment's operational results centered on three numbers:

• Up to 40% lower HVAC runtime, by matching ventilation to floor-level occupancy instead of a fixed timetable
• A 20% to 40% drop in demand-service mismatch, as cleaning and facility services shifted to follow real usage
• About 1,200 hours of cleaning labor saved per year, plus up to 15,000 gallons of water, through demand-based scheduling

Cutting HVAC runtime lowers the energy bill immediately, and it slows compressor wear, delaying the next capital replacement. The cleaning and water savings run on the same utilization sensors, replacing fixed routes with demand-based resource allocation and feeding a sustainability report with hard numbers.

Employee Experience Outcomes

How the workspace feels decides whether people come in at all in a hybrid office, and a stuffy or overcrowded floor is a reason to stay home.

Acting on the occupancy data, the team redistributed space and adjusted scheduling to ease peak occupancy. Peak crowding dropped by 15% to 30%. Peak congestion windows fell by 20% to 40%, smoothing the daily flow for people in the building.

The sensing stays anonymous, with no cameras, no device data, and no PII. The sensor-selection criteria below cover privacy requirements in detail.

The Multi-Use Case ROI

A sensor bought only to trim HVAC has to earn back its cost on energy savings alone. The math is tight, which is how a promising pilot can stall before it scales.

The same sensors here fed cleaning, space planning, and congestion management at the same time. So the cost spread across several budgets instead of one, and the case for scaling no longer rested on the energy line.

Space optimization is the clearest example. Over time, the same feed becomes workplace utilization data, tracking occupancy rates, peak usage times, and longer-term trends. The same data flags unused space, underused space, and floors running near total capacity.

Facilities management reads those metrics to redraw layouts and resize meeting rooms, workstations, and collaborative spaces around real room usage. Hot-desking ratios set from the data give teams flexibility as attendance shifts, and the payoff is less square footage to lease.

HVAC savings may not even be the biggest financial return. Real estate costs usually trail only payroll, so the space usage insights riding on these sensors can outweigh the energy reduction that justified them.

Butlr's thermal sensors power this kind of multi-use deployment. The sensor-selection criteria below explain what makes that possible.

Not All Sensors Deliver HVAC-Ready Data

Those outcomes hold only if the sensor produces data the BMS can use in real time. Plenty of sensors report occupancy without meeting that test. This can be an expensive mistake, as the sensor decision is a multiyear commitment that's hard to reverse once it's wired into a building.

Headcount, Not Just Presence

HVAC optimization depends on the number of people in a space, not just whether someone is there. DCV and zone-level setpoints both adjust to the headcount, so a sensor that only reports occupied or empty can't drive them.

Many buildings already have motion sensors for lighting and assume the same hardware will work for HVAC. It won't. A motion sensor can switch a system on, but it can't tell that system to run at half output for a half-full floor, which is where the savings come from.

Privacy That Passes Legal Review

Getting a sensor installed in an enterprise means clearing IT, legal, and, in Europe, a works council. Camera and Wi-Fi systems collect identifiable data, so they often stall in that review or get rejected after a pilot.

Thermal sensors collect no PII, so legal has little to review and approval comes faster. Faster approval matters most in large rollouts, where one works council can otherwise block deployment across every building.

Open Data Output

For HVAC to respond on its own, the occupancy count has to reach the BMS automatically. A sensor that keeps its data in a closed dashboard can show a facilities manager what's happening, but the system can't act on it unless someone reads the screen and adjusts setpoints by hand.

A manual step defeats the purpose of real-time control. Look for open APIs and webhooks that feed the count straight into the BMS, an integrated workplace management system (IWMS), or a business intelligence tool, so the sensor plugs into the systems a building already runs instead of becoming one more screen to watch.

The table below compares common sensor types on those three requirements.

Butlr's thermal sensors deliver anonymous, real-time occupancy data that feeds directly into BMS and IWMS platforms through open APIs. To see how this could look in your environment, let's talk.