Workplace occupancy detection is becoming increasingly important in large Activity Based Work (ABW) environments as it helps building and office management understand the utilisation and potential benefits of shared workplace. However, existing sensor-based technologies detect workstation occupancy in indoor spaces require extensive installation of hardware and maintenance incurring ongoing costs. Moreover, accuracy can depend on the specific seating styles of workers since the sensors are usually placed under the table or overhead. In this research, we provide a robust system called OccuSpace to predict occupancy of different atomic zones in large ABW environments. Unlike fixed sensors, OccuSpace uses statistical features engineered from Received Signal Strength Indicator (RSSI) of Bluetooth card beacons carried by workers while they are within the ABW environment. These features are used to train state-of-the-art machine learning algorithms for prediction task. We setup the experiment by deploying our system in a realworld open office environment. The experimental results show that OccuSpace is able to achieve a high accuracy for workplace occupancy prediction.
@inproceedings{DBLP:conf/percom/RahamanPLSRCKRS19,
author = {Mohammad Saiedur Rahaman and
Harsh Pare and
Jonathan Liono and
Flora D. Salim and
Yongli Ren and
Jeffrey Chan and
Shaw Kudo and
Tim Rawling and
Alex Sinickas},
bibsource = {dblp computer science bibliography, https://dblp.org},
biburl = {https://dblp.org/rec/conf/percom/RahamanPLSRCKRS19.bib},
booktitle = {IEEE International Conference on Pervasive Computing and Communications
Workshops, PerCom Workshops 2019, Kyoto, Japan, March 11-15, 2019},
doi = {10.1109/PERCOMW.2019.8730762},
pages = {415--418},
publisher = {IEEE},
timestamp = {Fri, 27 Dec 2019 00:00:00 +0100},
title = {OccuSpace: Towards a Robust Occupancy Prediction System for Activity
Based Workplace},
url = {https://doi.org/10.1109/PERCOMW.2019.8730762},
year = {2019}
}
© 2021 Flora Salim - CRUISE Research Group.