The presence and occupancy virtual sensors enable you to understand how spaces in your building are used. These sensors have been made to understand long-term trends, not instant values. You can find this sensor at the bottom of the device page with other virtual sensors. The data is only shown when the opening hours for the building are enabled.
Presence shows how long a room has been used for. The graph is updated every five minutes, indicating if the room was in use or not. This is then aggregated into total minutes in use.
By entering room height and area in device settings, occupancy will be available with the presence data as range columns. This represents the estimated number of occupants in the room.
It should be noted that buildings with natural ventilation will have a lower accuracy due to lack of regularity in the ventilation rates.
The graphs shown update every 5 minutes when the room is actively used, and every hour in quiet periods. Historical data will also be updated a few hours back based on more recent data.
The underlying algorithm uses all sensors and some metadata to calculate presence and occupancy. It analyses the room over long periods of time to improve the virtual sensor’s performance. The algorithm also needs a minimum of 8 hours of data without gaps to show the virtual sensors.
The accuracy of the data depends on multiple variables, including:
Size vs number of people in the room
Ventilation system and temperature gradients (mixing of air)
As a result, the algorithm will work best for rooms where the air is mixed evenly for the device to register readings appropriately. Rooms with uneven mixing, such as meeting rooms with open doors or corridors will have less accurate results.
The data will only show if someone was present in the space and is not sufficient for identifying individuals. These sensors will not give real-time presence data such as signaling if seats are in use or not. The sensor is also unable to say if a room is at full capacity.