Canonical URL: https://density.io/resources/continuous-office-occupancy-measurement # How should you choose sensors for continuous office occupancy measurement? Continuous office measurement needs usable data, not just a device that stays switched on. The system must measure the right thing, maintain its connection and make interruptions visible before they distort a space decision. [Density Waffle](https://density.io/waffle) provides anonymous radar measurement for rooms and phone booths, with plug-in power and real-time and historical data. Evaluate the sensor output, connection and reporting together before expanding the rollout. ## Define the job before choosing the sensor An occupancy switch that controls lights answers a different buying question from a system measuring how rooms are used over weeks. If your goal is continuous workplace measurement, specify the occupancy output and historical reporting you need before comparing hardware. Waffle reports **zero, one, two or three-plus people**. Those categories support questions about availability and individual versus group use. Counts of three or more are grouped into a single category. [Waffle’s published specifications](https://density.io/waffle) make that boundary explicit. If your decision requires exact larger headcounts, ask Density to propose the appropriate configuration and validation method. ## Remove battery servicing from the room plan Waffle uses plug-in power through a standard outlet and USB-C cable, according to [Density’s setup documentation](https://support.density.io/hc/en-us/articles/37483423312923-How-does-Waffle-work). There are no sensor batteries to replace. The product page specifies dual-band Wi-Fi and WPA2 PSK. Where outlets and suitable Wi-Fi are available, Waffle avoids battery servicing. Identify the outlet, secure the cable and confirm network approval. Assign someone to review power or Wi-Fi changes, including those caused by a furniture move. ## Keep missing data separate from empty rooms Density’s [API reference](https://developers.density.io/docs/) documents health-aware presence, current and historical sensor health, and an `up_time` metric describing the share of a reporting bucket when a sensor was online and able to provide data. These give an integration a documented basis for identifying availability problems. Use those signals in your reporting rules. An empty room and an unavailable measurement should have different meanings. Density’s [health guidance](https://developers.density.io/understand_health/) recommends ignoring live data from spaces marked offline or unknown and treating degraded spaces cautiously. Use health signals to assess availability and completeness. Validate sensor placement and count output alongside them. ## See what an interruption does to a utilization report Suppose a room has eight scheduled operating hours. In this invented example, six hours have usable observations: the room is occupied for three hours and empty for three. The remaining two hours have no usable data. Assume presence is known throughout each observed period; this is an illustration, not a Density performance result. | Reporting choice | Calculation | What it tells you | | --- | --- | --- | | Treat missing time as empty | 3 occupied hours ÷ 8 scheduled hours = 37.5% | Assumes the room was empty during the interruption | | Use observed time only | 3 occupied hours ÷ 6 observed hours = 50% | Describes use during the periods with usable observations | | Report observation coverage | 6 observed hours ÷ 8 scheduled hours = 75% | Shows how much of the scheduled day the report can describe | The observed occupied-time rate is 50%, with 75% observation coverage. That does not establish a 50% rate for the whole day. If the room was empty throughout the two missing hours, its full-day occupied-time rate would be 37.5%. If it was occupied throughout them, the rate would be 62.5%. Without further evidence, the missing period cannot be assigned to either outcome. This range describes uncertainty from missing time under the example's assumptions. It does not account for errors in the observed readings. Check measurement validity separately from availability, and agree with Density how the proposed report will show incomplete periods before using it to compare rooms or reduce space. ## Check the installed system before expanding Use a small rollout to agree what continuous measurement means for your office: 1. **Define the output.** Record each room boundary, required occupancy categories, operating hours and reporting interval. 2. **Check real use.** Observe empty periods, seated individual work and group meetings. Compare observations with the reported categories, including three-plus. 3. **Check continuity.** With IT approval, rehearse a brief connection interruption in a test space. Verify that the reporting workflow distinguishes unavailable data from vacancy and shows recovery. 4. **Agree reporting rules.** Set an acceptable data-availability threshold for your use case. Decide how incomplete intervals appear in charts and comparisons. 5. **Assign maintenance.** Name owners for connectivity, power, placement changes, incident review and support escalation. Use these checks to define acceptance for your deployment. Ask for product-specific recovery behavior and support terms before committing to a rollout. Bring your room list and continuity requirements to [a Density demo](https://demo.density.io), or email [sales@density.io](mailto:sales@density.io). Ask how the proposed deployment reports room use, identifies interruptions and supports your reporting rules. The [occupancy API guide](https://density.io/resources/real-time-occupancy-api) covers health-aware applications. The [pilot scorecard](https://density.io/resources/occupancy-sensor-pilot-scorecard) shows why missing observations belong in a separate coverage measure.