Beyond presence: addressing passive occupancy in desk utilisation metrics
Kiia Aurora Einola, Pasi TakalaPurpose
Measuring desk-level occupancy in offices using sensors provides insights into space utilisation. However, presence sensor data alone cannot distinguish between actively used desks and those passively occupied because of personal belongings left behind. This study aims to evaluate how incorporating a clean desk policy (CDP) time delay (TD) filter in desk occupancy data affects the accuracy of desk availability assessments.
Design/methodology/approach
Desk occupancy was monitored using three types of sensors across ten desks, with ground truth data collected through observational walking tours. A total of 5,315 observations and corresponding sensor statuses were analysed to assess the impact of applying CDP TD filters.
Findings
Applying a CDP TD filter improved the accuracy of sensor-based desk availability measurements. Implementing a 60-min filter for desk-level increased accuracy by 22%, while a 120-min filter improved accuracy by 27% compared to unfiltered data.
Research limitations/implications
This study tested only two types of sensors, and broader generalisability would require additional sensor technologies and varied office environments.
Practical implications
Facility managers relying on sensor data without accounting for passive occupancy may misinterpret desk availability, leading to an estimated 30% overstatement of available workspaces. Integrating CDP TD filters can enhance decision-making in workspace management and optimisation.
Originality/value
Previous studies evaluating the accuracy of desk-level sensor measurements do not account for passive occupancy. This study incorporates this third occupancy state into desk-availability assessment.