This chapter explains how Spacewell calculates various occupancy metrics.
Line of thought
Occupancy and headcount values are often misinterpreted by data analysts.
This page explains the Spacewell viewpoint on this subject matter.
Business case
This business case is only considering presence sensors in combination with hourly timeslots.
Let’s assume a simple floor 2 with this layout:
- Room ABC contains 10 desks (WP1 to WP10)
- In an open space we have another 12 desks (WP11 to WP20 and WP21 and WP22)
- 4 meeting spaces (MR1 to MR4)
- 3 focus spaces (FS1 to FS3)
On a given day , let’s look at how these spaces are used per hour:
- green cells = average headcount detected by presence sensors
“headcount” means the average value per hour, for 1 specific space
- grey cells = no presence detected
Info
The headcount values per hour are not always 0 or 1. During any timeslot, sensors may detect presence or absence within minutes.
Consequently, Spacewell always registers decimal headcount values per timeslot:
- headcount = average value of all sensor readings during the timeslot interval
(as shown in the image above)
Likewise, the occupancy rate per timeslot = average headcount (presence) / nr of monitored spaces. This is the most accurate and detailed metric for space occupancy.
As illustrated above, the occupancy rate changes during each business day. The example shows values ranging between 4% and 46% on a single day!
What happens when aggregating hourly values to daily?
In our business case, it leads to these values:
The daily headcount is 6,50 on average, with a peak of 13,25 and a low of 1,25.
Accordingly, the daily occupancy rate is 22% on average, with a peak of 46% and a low of 4%.
Info
When observing a full business day, this example illustrates the huge difference between average and peak values! Peak values represent a moment in time throughout the day where sensors registered the highest headcount and occupancy rates.
When observing a long period of time (multiple days) via timeline charts in dashboards, these definitions matter:
Term | Definition | Comment |
|---|---|---|
Avg occupancy (%) | Is technically a triple aggregation:
| Gives insight in average occupancy on a global level |
Max occupancy (%) | Is technically a triple aggregation:
| Gives insight in highest average occupancy levels. Not to be confused with peak occupancy! |
Peak occupancy (%) | Is technically a triple aggregation:
| Gives insight in the average peak occupancy levels. |