Space Utilization Sensors: What to Know Before You Buy

The space utilization technology you choose determines which metrics you get (headcount vs. binary presence, dwell time vs. motion detection), which spaces you can and can't cover, and how quickly you can move from pilot to portfolio-wide data.
‍
After deploying over 20,000 sensors across more than 40 million square feet, we've seen the same questions come up in nearly every engagement: 

• Which floors can be consolidated without displacing teams that rely on them? 
• Are 10-person rooms consistently occupied by groups of two? 
• Is Wednesday's low badge count a policy problem or a measurement problem?

The answers depend on collecting the right data, which starts with choosing the right sensor. This guide covers the technologies available, the metrics each one can and can't deliver, and the deployment decisions that determine whether you end up with actionable data or noise.
‍

What Space Utilization Means (and Why It's Not the Same as Occupancy)

Many vendors use occupancy and space utilization interchangeably, but they describe different things. Occupancy is a point-in-time measurement that indicates how many people are in a space right now. Space utilization builds on workplace occupancy data to reveal patterns, trends, and actionable insights over time.
‍

The Metrics That Define Space Utilization

‍‍

Sensor Technologies for Space Utilization (and The Data They Provide)

Every sensor technology has tradeoffs. The best fit occupancy sensors for your portfolio depend on which utilization metrics you need, where you need to deploy, and how fast you need to scale. The technology you choose determines whether you get real-time data and reliable usage patterns or a partial picture with gaps.

‍

‍

Thermal Sensors

Thermal sensors detect body heat signatures using an array of thermal sensing elements. They produce a low-resolution heat map that distinguishes individual occupants without capturing any visual information.
‍
They don't capture images or personally identifiable information (PII), and they don't have cameras. The data output is anonymous by design (not by policy) which means there's no PII to encrypt, store, or accidentally expose.
‍

Utilization metrics they deliver:

• Accurate headcount, including density versus capacity analysis
• Reliable presence detection for stationary occupants, including dwell time and utilization rate data
• Traffic counting when deployed in doorway or corridor mode, supporting turnover and flow pattern analysis
• Full coverage in all lighting conditions, including spaces where cameras can't go, such as restrooms, prayer rooms, wellness spaces, and healthcare settings

Where they have limits

Thermal sensors offer lower resolution than camera or LiDAR systems for very granular movement path analysis across large open floor plates. If you need to understand exactly how individuals navigate a 50,000-square-foot trading floor, a thermal sensor won't give you that level of detail. For most enterprise use cases, room-level and zone-level analytics are sufficient for data-driven decisions.

Deployment timeline

Battery-powered thermal sensors don't require an electrician or wired infrastructure. Installation is comparable to mounting a smoke detector. A team can deploy hundreds of sensors overnight, and the sensors begin capturing data as soon as they're online. Typical time from installation to actionable insights is three to four weeks, including a tuning period.

Best for

These sensors are ideal for portfolio-wide deployments where privacy compliance, deployment speed, and scalability are priorities. If you need data next month instead of next quarter, and you need it across every space type in your portfolio, thermal sensors are the practical choice.

Butlr's thermal sensors deliver 95% headcount accuracy with zero PII collection and deploy without wiring or electricians. Most teams go from installation to portfolio-wide data in under a month. Request a demo and we'll walk through what deployment looks like for your building count and floor mix.

Camera-Based / Computer Vision

Camera-based systems use vision cameras with AI-powered image recognition to detect, count, and sometimes follow occupants through a space. The accuracy of these systems has improved in recent years, but the technology introduces constraints that go beyond the hardware itself.
‍
Utilization metrics they deliver:

• High-accuracy headcount in large open areas
• Rich behavioral data including direction of movement, queue length, and movement paths
• Dwell time and density analysis

Where they have limits

Cameras can't be deployed in privacy-sensitive spaces like restrooms, wellness rooms, or healthcare environments. This creates permanent blind spots in portfolio-wide utilization data. Even if you cover 90% of your floor area, the 10% you can't cover may include some of the most important spaces for understanding how a building functions.

These privacy issues create substantial drawbacks for most enterprise deployments. Legal review, works council approval, and employee pushback can slow or block rollouts entirely.

Some vendors offer a blurred vision camera option to reduce identifiability. But even with blurring or edge processing, the hardware is physically capable of capturing identifiable images. This creates ongoing compliance risk regardless of software settings.

Deployment timeline

Camera-based sensors take weeks to months to deploy per building, depending on legal review and infrastructure requirements. Wired installation typically requires an electrician.

Best for

These sensors are best in high-traffic, lower-privacy-sensitivity locations like lobbies, cafeterias, and large office spaces where rich behavioral data adds value. They're difficult to scale across full portfolios due to privacy constraints and installation complexity.
‍

Passive Infrared (PIR)

PIR sensors detect changes in infrared radiation caused by movement, functioning as basic motion detectors. They're among the most common infrared sensors in commercial buildings, primarily because they've been used for decades to trigger lighting and HVAC systems. But motion detection and occupancy detection are very different capabilities.

Utilization metrics they deliver:

• Binary presence/absence, which is equivalent to a basic utilization rate

Utilization metrics they can't deliver

PIR sensors can't count people. To these sensors, one person in a conference room looks the same as five do.
‍
They also can't reliably detect stationary occupants. A person sitting still for more than a few minutes essentially disappears from the sensor's perspective. This makes dwell time data unreliable.
‍
In larger spaces, they're prone to false negatives. In addition, HVAC airflow can trigger false positives.

Deployment timeline

PIR sensors are Quick to install at low cost per unit. But the limited data value often means you'll need to redeploy a more capable technology later. This means you effectively end up paying twice.

Best for

These sensors are best for simple lighting or HVAC triggers. They aren't sufficient as a primary source of space utilization data. In some cases, they're worth considering alongside more capable sensors as a low-cost supplementary layer.

LiDAR

Light detection and ranging (LiDAR) sensors use laser pulses to create 3D point cloud maps, detecting people by their physical shape. They represent the high end of spatial accuracy among workspace occupancy sensors, and they can resolve individual positions within centimeters.

Utilization metrics they deliver:

• Very high-accuracy headcount and spatial positioning
• Detailed traffic pattern and zone utilization data, including precise movement paths across open areas
• Environmental mapping of how people distribute across zones, useful for optimizing floor layouts
• Reliable performance in all lighting conditions

Where they have limits

That precision is expensive. LiDAR sensors cost several times more per unit than thermal or PIR alternatives. Plus, each unit requires professional installation, ceiling mounting at controlled heights, and calibration to the geometry of the room. Moving a LiDAR sensor to a different floor or building often means recalibrating from scratch.
‍
Coverage area per sensor is also limited. Large open floors may need multiple units to avoid dead zones, which drives the cost higher.
‍
And while LiDAR point clouds are less identifiable than camera footage, they can be detailed enough to reconstruct body shapes and gait patterns. Some European data protection authorities have flagged this as a potential concern, so legal review is still part of the deployment process.

Deployment timeline

LiDAR sensor installation takes weeks to months per location, due to installation complexity and calibration. Adding new buildings requires repeating the full setup process, which makes rapid multi-building rollouts impractical.

Best for

They're ideal for high-value single locations like flagship offices, innovation labs, or retail environments where extremely granular spatial data justifies the cost and timeline. But they're rarely practical for portfolio-wide deployments where dozens or hundreds of buildings need coverage.

Wi-Fi / Bluetooth Tracking

Wi-Fi and Bluetooth Low Energy (BLE) systems detect wireless signals from personal devices to estimate occupancy and location. Wi-Fi tracking estimates a device’s location by measuring signal strengths and scan data from multiple nearby access points. BLE tracking uses short-range beacons that broadcast identifiers, allowing nearby devices to estimate proximity based on signal strength.

Utilization metrics they deliver:

• Rough zone-level traffic patterns using existing infrastructure

Where they have limits

For this sensor type, accurate headcount is the core problem. Not all occupants carry detectable devices, some carry multiple devices, and MAC address randomization makes consistent counting unreliable.
‍
Location accuracy typically varies by anywhere from three to 10 meters, which rules out room-level utilization. And in any space where occupants don't bring devices, the data goes dark entirely.

Deployment timeline

These sensors can leverage existing Wi-Fi infrastructure, which makes initial deployment fast. But accuracy limitations often mean supplementing with dedicated sensors, which adds cost and complexity.

Best for

They're helpful for providing directional, macro-level data that supplements other sources. But they aren't reliable as a primary space utilization sensor.

‍

Turning Data Into Decisions: Common Use Cases

Here are five ways space utilization sensors inform decision-making, turning data into action.

Portfolio Right-Sizing

Comparing utilization rate and peak utilization versus average occupancy data across buildings over months helps you identify which leases to renew, renegotiate, or exit. For instance, a building running at 35% average utilization for six months straight tells a different story than one that spikes to 90% on Tuesdays and Thursdays.

What you need: Headcount-capable workplace sensors deployed across all buildings in scope, not a sample. Making right-sizing decisions based on partial data carries serious financial risk.

Hybrid Work Policy Validation

Day-of-week utilization patterns show whether hybrid work policies are performing as intended or whether certain days are dramatically over- or under-occupied. If your workplace is empty on Wednesdays and overflowing on Tuesdays, you may have a collaboration scheduling problem, not a space problem.

This is also where badge data falls short. A Wednesday with low badge swipes could mean the policy isn't working, or it could mean people are entering behind others. Sensor data tells you what's happening across the floor and the collaborative zones, not just at the door.

What you need: Presence detection that covers full floors, not just entry points.

Conference Room Optimization

Comparing density versus capacity data reveals which rooms are over-provisioned and which are undersized. For example, a 20-person boardroom that consistently hosts groups of three or four is consuming premium square footage for a use case that a six-person huddle space could serve.

Ghost booking data takes this further. It identifies rooms that are reserved but never occupied or those that are occupied for only the first 15 minutes of an hour-long booking.

In the absence of room-level sensors, both problems remain invisible. Booking system data alone shows reservations, not usage. And because ghost bookings block availability for other teams, the problem grows quickly. When people can't find rooms, they might over-book as a hedge, which inflates reservation rates and makes the data even less reliable.

What you need: Room-level headcount sensors in every bookable room, integrated with your booking system to compare reservations against observed occupancy.

Demand-Based Operations

Real-time occupancy data can shift cleaning and HVAC operations from fixed schedules to usage-based triggers. Instead of cleaning every floor every night, facilities teams can prioritize floors that were occupied and skip floors that weren't.
‍
Instead of conditioning an entire building to the same temperature, smart building automation systems can adjust environmental conditions floor by floor. This reduces heating or cooling on unoccupied areas and ramps up only when sensors detect arrivals.
‍
This change can meaningfully reduce operating costs. Cleaning labor is one of the largest controllable resources in facilities budgets, and HVAC accounts for a major share of commercial building utility and energy costs.
‍
Shifting to a demand-based model improves energy efficiency without reducing service quality, because you're still cleaning and conditioning every space that's used. The result is a more productive workplace with lower operating costs.

What you need: Sensors with REST API and webhook integration into your existing BMS and computerized maintenance management system (CMMS) platforms. The sensor platform should push data to your operations tools in real time, not require manual export or batch uploads.

ESG and Sustainability Reporting

Most buildings report energy efficiency on a per-square-foot basis. This can make a half-empty building look efficient when it's anything but. Occupancy-normalized energy data, measured as energy per occupant rather than per square foot, gives a more honest picture of how efficiently a building serves the people inside it.
‍
This distinction matters for sustainability reporting. For instance, a building running at 20% occupancy five days a week may meet its per-square-foot energy targets while consuming far more energy per person than a fully occupied building with higher total consumption. Without headcount data, though, you can't make this distinction. As a result, your environmental, social, and governance (ESG) reports will reflect the building's design performance rather than its operational reality.

What you need: Consistent, accurate headcount data paired with energy systems integration. The headcount data needs to be granular enough (floor-level at minimum) to map against energy metering zones.

Common Mistakes When Selecting Space Utilization Sensors

As you compare and consider space utilization sensors, watch for these miscalculations.

Treating Badge Data as Utilization Data

Badge scanners tell you who entered the building. But they don't tell you where those people went, how long they stayed, or whether they used the spaces they booked. Organizations that rely on badge data for utilization decisions consistently overestimate actual space usage.

‍

Choosing a Technology That Can't Cover Your Full Portfolio

If your sensor can't go in sensitive environments, you end up with permanent blind spots. If it requires wired installation, scaling across 50 buildings becomes a multiyear project. The technology you choose for a pilot needs to be the technology that works at full scale.

‍

Optimizing for Sensor Cost Instead of Data Quality

The cheapest sensor per unit (PIR) is also the one that produces the least useful utilization data. Instead, it's much more useful to compare total cost of ownership (TCO) at scale.

Consider the hardware, installation labor, ongoing maintenance, network infrastructure, and the cost of decisions made on incomplete data. When evaluating vendors, ask for a TCO estimate that includes all five of those categories across your full portfolio, not just unit price for a pilot.

‍

Deploying Sensors Without Defining the Decisions They Need to Inform

A vague goal like wanting occupancy data doesn't tell you enough to choose a sensor, a deployment scope, or a timeline. Are you measuring meeting room usage to optimize your booking system? Deploying desk sensors to validate an office layout change?

A useful requirements statement ties the data to a decision. For example: room-level headcount across 12 buildings, ready in time for a Q3 lease portfolio review. Start with the decision, then work backward to the metrics, the technology, and the deployment plan.

‍

Ignoring the Integration Layer

A sensor platform that doesn't work with your existing integrated workplace management system (IWMS), BMS, or BI tools creates a data silo. Utilization data is only as valuable as the systems it connects to. Prioritize platforms with open APIs and proven integrations over those with proprietary dashboards.

‍

Questions to Ask Before You Commit

These are the questions that separate a productive vendor conversation from a demo that looks good but doesn't address your requirements.

‍

• What utilization metrics does this sensor produce? Get specifics like headcount, presence, dwell time, and traffic. If the vendor answers with vague terms like "occupancy intelligence," push for which of these core metrics the hardware can deliver and at what accuracy.
• Can this technology be deployed in all space types across our portfolio, including privacy-sensitive areas? If the answer involves carve-outs for restrooms, wellness rooms, or healthcare environments, you'll have permanent gaps in your data.
• What does a realistic deployment timeline look like for our building count? Ask for weeks or quarters, and ask what's included. Some vendors quote installation time without accounting for legal review, works council approval, or network provisioning.
• What's the TCO at portfolio scale? Unit price is one input. Installation labor, electrician requirements, network infrastructure, ongoing maintenance, and battery or hardware replacement cycles all factor into the real number. Ask for a TCO estimate across your full portfolio, not just the pilot.
• How does the data get into our existing systems? Look for open APIs and webhooks that push data into your IWMS, BMS, and BI tools. If the answer is a proprietary dashboard with manual CSV export, you're building a data silo.
• What happens if we need to move sensors between floors or buildings as our space plan changes? Your space plan will change. Sensors that require recalibration, rewiring, or professional reinstallation every time you reorganize a floor add friction and cost to every future adjustment.

‍

Butlr's thermal sensors and API-first platform check every box on that list: 95% headcount accuracy, full privacy compliance, battery-powered installation, and open API integration with your existing BMS, IWMS, and BI tools. Request a demo with your building count and we'll put together a deployment timeline and TCO estimate for your portfolio.