Abstract
Many high-consuming electricity loads in retail stores are currently highly intertwined with human activities. Without knowledge of such activities, it is difficult to improve the energy efficiency of the loads operation for sustainability and cost reasons.
The increasing availability of Internet of Things sensors and devices promise to deliver rich data about human activities and control of loads. However, existing proposals for building operating systems that should combine such data and control opportunities do not provide concepts and support for activity data. In this paper, we propose an Activity-Tracking Service (ATS) for building operating systems.
The service is designed to consider the security, privacy, integration, extendability and scalability challenges in the building setting. We provide initial findings for testing the system in a proof-of-concept evaluation using a set of common Internet of Things sensors and devices.
Key idea
Today’s BOS proposals (BOSS, XBOS, BRICK, Bosswave, Building Depot 2.0) model what is in the building but not what is happening in it. The ATS adds an activity layer on top – with explicit modelling and labelling of activities like “cleaning the floors”, “pre-heating ovens”, “refilling refrigeration units” – so that DR applications can reason about why a load is on, not just that it is on.
Why this matters
A naïve DR strategy might shift refrigeration cycles to off-peak periods. But if the cycle was triggered by just-restocked warmer products, deferring the cycle damages food quality. Activity awareness lets the BOS distinguish “the freezer is on because of its schedule” from “the freezer is on because of restocking” – and treat them differently in DR decisions.
Citation
Hviid, J., & Kjærgaard, M. B. (2018). Activity-Tracking Service for Building Operating Systems. In 2018 IEEE International Conference on Pervasive Computing and Communications Workshops: PerCom Workshops, pp. 854–859. IEEE Press. https://doi.org/10.1109/PERCOMW.2018.8480362