OccuRE: An Occupancy REasoning Platform for Occupancy-Driven Applications

Abstract

Occupant behavior determines a large share of the energy consumption of buildings. Software applications driven by information about occupant behavior provide a mean to optimize this share. However, existing systems for sensing occupancy behavior provide technology-specific APIs statically coupled to the type of computed occupancy information. Software platforms for developing applications for buildings do also not provide abstractions for occupancy behavior. Therefore, technology lock in and lack of proper abstractions wreck the development of occupancy-driven applications. In this paper we present the design, implementation and evaluation of OccuRE, a stream-based Occupancy REasoning platform. OccuRE provides a technology agnostic API for accessing occupancy information to significantly improve portability. The platform uses a component-based computation model with dynamic composition to calculate and reason about occupancy behavior. Together these elements avoid that developers need to deal with technology-specific processing of sensor data to ease application development. Through microbenchmarks we show that OccuRE successfully and efficiently computes occupancy information for technology-heterogeneous building instrumentations. We use the development of three prototype applications to demonstrate that the API of OccuRE (i) enables several types of occupancy-driven applications, (ii) that the applications-by using the interface-achieve portability in regards to occupancy information computation and (iii) that the application code avoids handling sensor data processing.

DCount - A Probabilistic Algorithm for Accurately Disaggregating Building Occupant Counts into Room Counts

“…al. [5], compared count data obtained from PIR sensors in a building with ground-truth data. The results highlight that PIR sensors are not fit for occupant counting because of a RMSE of 21.7.…”

“…However, often these modalities alone only provide information with a high Root Mean Squared Error (RMSE). For instance, Kjaergaard et al [5] report a RMSE of 21.7 for counting occupants using PIR sensors in a small office building and [6] report an accuracy around 50% for counting occupants in a large hospital complex using WiFi access points. For CO 2 sensors existing work [7], [8] has shown that such sensors are often too error-prone to use for occupant sensing.…”

Demand response with model predictive comfort compliance in an office building

“…For the DR analyzer / scheduler we use the Controleum Framework [2], a multi-objective optimization framework using genetic algorithms. Zone-based indoor climate models are developed in Modelica, electricity forecasting models [9] in Java and wrapped in FMU wrappers, and occupancy prediction is done by the OccuRE system [10]. External data is accessed via drivers developed in Python and stored in sMAP [11] which offers fast querying over a timeseries database.…”

PLCount

“…One line of work has studied reusing common building sensors for occupancy counting including CO2 sensors, PIR sensors, energy metering, sensors of HVAC systems or WiFi access points [2] but often these sensors only provide counts with a high Root Mean Squared Error (RMSE). For instance Kjaergaard et al [8] report a RMSE of 21.7 for PIR in a small office building. Beltran et al [1] explore the idea of densely deploying lightweight thermal sensors for occupancy counting in all areas of a building.…”

“…3D camera-based counting sensors are quite accurate in the short term. For instance, Kjaergaard et al [8] report a RMSE of 3.3 for a three hours evaluation. However, particular detection problems associated with 3D cameras includes occlusion, pixel intensity fluctuations, and poor lighting conditions resulting in false positive and false negative counts.…”