Occupancy level prediction based on a sensor-detected dataset in a co-working space

Abstract

Hybrid working has reshaped people's routines and working habits, while the workplace needs to evolve with the new working pattern. Co-working space is seen as an alternative work environment, for cost-effectiveness, the opportunity for flexible design and multi-use. This study investigates the occupancy patterns and occupants' behaviour using multiple occupancy sensor data with a twelvemonths sample. Data-driven AutoRegressive Integrated Moving Average (ARIMA) time series model is applied to predict office occupancy in a co-working space in London. The results reveal some spatial-temporal variations in the number of occupants based on the detected locations. The spatial distribution of occupants around different working areas in the co-working space is plotted to demonstrate the seat preferences and its temporal occupancy density variation.