Workplace Occupancy Sensors: Types, Accuracy and How to Choose

Workplace occupancy sensors measure how many people use an office floor, meeting room or desk, and when, so companies can size their real estate, fix meeting room shortages and run buildings on actual use instead of guesses. The right sensor depends on the question: entrance counters give floor headcount, room sensors show meeting room use, and desk sensors show desk utilization.

Hybrid work has made these questions urgent. Many offices are crowded on Tuesday and empty on Friday, meeting rooms are booked but not used, and real estate teams need evidence before they consolidate floors or redesign space. This guide compares the main types of workplace occupancy sensors, explains accuracy and privacy, and gives a planning checklist.

Three levels of workplace occupancy data

Level	Question it answers	Typical sensor
Building and floor	How many people are in the building or on each floor, hour by hour?	Overhead people counters at entrances, lifts lobbies and stair doors
Room and zone	Is the meeting room in use, and by how many people? How busy are kitchens, lounges and focus areas?	Room presence or counting sensors on the ceiling
Desk	Which desks are used, how often and for how long?	Under-desk sensors or overhead sensors covering several desks

Many projects start at floor level, because a handful of counters covers a whole building and already answers the biggest question: how much space is needed on a peak day. Room and desk sensors are then added where the decisions are made. For desk-level detail, see our desk occupancy sensor guide.

Types of workplace occupancy sensors

PIR motion sensors

Passive infrared (PIR) sensors detect movement of warm bodies. They are cheap, battery-powered and already common for lighting control. For occupancy analytics they answer only "in use or not", and a person sitting still at a desk or in a long meeting can be reported as absent. Most systems compensate with time-outs, which blur the data.

Thermal array sensors

Low-resolution thermal sensors on the ceiling see people as heat spots, including people sitting still. They can count a small number of people in a room or desk cluster and capture no identifiable images. Accuracy drops when people sit very close together or when heat sources such as laptops and radiators are in view.

Radar (mmWave) sensors

Millimeter-wave radar detects tiny movements such as breathing, so it registers seated people reliably and works through light materials. Radar sensors can estimate the number of people in a room and are privacy-friendly. They need careful tuning in small rooms with glass walls or moving objects such as fans.

3D stereo and time-of-flight counters

Overhead 3D counters at doorways count people walking in and out with high accuracy, even in groups, and are the standard choice for floor and building headcount. Mounted over open areas, they can also measure zone occupancy and dwell time. Because they count from depth data, they do not need to store images. See our comparison of infrared, camera and 3D people counters for the technology trade-offs.

Network and badge data

Wi-Fi connections and access badge swipes are often used as a free first estimate. They are useful for trends but biased: one person may carry several devices, visitors may not badge in, and people tailgate through doors. Most workplace teams use them alongside sensors rather than instead of sensors.

Comparison at a glance

Sensor	Counts people	Detects seated people	Privacy	Best use
PIR	No	Unreliable	No images	Lighting control, basic room in-use signal
Thermal array	Small groups	Yes	No identifiable images	Meeting rooms, desk clusters
Radar (mmWave)	Estimates	Yes	No images	Meeting rooms, focus rooms
3D stereo or ToF counter	Yes, accurately at doorways	In zones, yes	Depth data; images not required	Floor and building headcount, zones
Wi-Fi and badge data	Estimates	Indirectly	Personal data involved	Trends, combined with sensors

Metrics that matter for workplace teams

• Peak occupancy per floor. The busiest hour on the busiest day sets how much space and how many desks you actually need.
• Average desk utilization. The share of desks in use during working hours, often far below what booking data suggests.
• Meeting room utilization and right-sizing. How often rooms are used, and how many people use them. Large rooms used by one or two people point to a need for more small rooms.
• No-shows and ghost bookings. Rooms booked in the calendar but empty in reality. Connecting sensors to the booking system can release them automatically.
• Weekday patterns. Attendance by weekday guides hybrid policies, cleaning and catering schedules.

Privacy and employee trust

Occupancy sensing works best when employees trust it. Choose sensors that measure space use, not individuals, and avoid storing images. Explain to staff what is measured, why, and who sees the data, and involve employee representatives or works councils early where they exist. Aggregate data by zone and time, and set short retention periods for raw data.

Integration: where the data should go

Occupancy data becomes valuable when it reaches the systems that act on it:

• Workplace analytics dashboards for real estate and facilities teams.
• Room booking systems, to release no-show bookings and show free rooms on panels and apps.
• Building management systems, so heating, cooling, ventilation and lighting follow real occupancy.
• Cleaning and catering schedules, based on actual use of kitchens, washrooms and meeting rooms.

Check that sensors offer an open API, MQTT or a documented export. Our people counting API guide shows typical data formats.

Planning checklist

1. Write down the decisions the data should support: consolidating floors, adding meeting rooms, changing desk ratios, saving energy.
2. Choose the level of detail for each area: floor, room or desk.
3. Run a pilot on one floor for four to eight weeks before rolling out.
4. Check power and connectivity: PoE, battery life, wireless coverage and IT security approval.
5. Agree privacy rules and communication with employees before installation.
6. Validate counts against a manual check at a few entrances and rooms.
7. Assign an owner who reviews the data monthly and turns it into decisions.

Where Pyroglaux fits

Pyroglaux supplies 3D stereo people counters for floor and building headcount, such as the PCBot-R40, which also measures zone occupancy and dwell time, alongside the wider smart counting range. Send us your floor plans, entrance list and the questions you need to answer, and we will propose a sensor layout for a pilot floor.

Frequently Asked Questions

What is the difference between occupancy and utilization?

Occupancy is how many people are in a space at a given moment. Utilization compares that use with capacity over time, for example the share of working hours a meeting room is in use, or the average share of desks occupied on a floor.

Are PIR sensors good enough for workplace occupancy?

PIR sensors are fine for knowing whether a room or desk is in use, as long as people move. They cannot count people and can report a room as empty when someone sits very still. Use thermal, radar or 3D sensors when you need head counts or reliable presence for seated people.

Do workplace occupancy sensors record employees?

Most workplace occupancy sensors are designed not to: PIR, radar and low-resolution thermal sensors capture no identifiable images, and 3D counters can count from depth data without storing images. Involve your works council or employee representatives early and publish what is measured and why.

How many sensors does an office floor need?

It depends on what you measure. Floor-level headcount needs counters at each entrance to the floor. Meeting room use needs one sensor per room. Desk-level data needs a sensor per desk or an overhead sensor covering a group of desks.

Can occupancy sensors reduce energy costs?

Yes, when the data is connected to the building management system. Heating, cooling, ventilation and lighting can follow real occupancy instead of fixed schedules, and unused floors can be closed on quiet days.