Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Gap in Time: The Challenge of Processing Heterogeneous IoT Data in Digitalized Buildings

Xiachong Lin, Arian Prabowo, Imran Razzak, Hao Xue, Matthew Amos, Sam Behrens, Flora D. Salim

TL;DR

This study investigates the diverse dimensions of IoT data heterogeneity in both intrabuilding and interbuilding contexts, examining their implications for predictive modeling and advocates collaborative efforts to establish high-quality public datasets.

Abstract

The increasing demand for sustainable energy solutions has driven the integration of digitalized buildings into the power grid, leveraging Internet-of-Things (IoT) technologies to enhance energy efficiency and operational performance. Despite their potential, effectively utilizing IoT point data within deep-learning frameworks presents significant challenges, primarily due to its inherent heterogeneity. This study investigates the diverse dimensions of IoT data heterogeneity in both intra-building and inter-building contexts, examining their implications for predictive modeling. A benchmarking analysis of state-of-the-art time series models highlights their performance on this complex dataset. The results emphasize the critical need for multi-modal data integration, domain-informed modeling, and automated data engineering pipelines. Additionally, the study advocates for collaborative efforts to establish high-quality public datasets, which are essential for advancing intelligent and sustainable energy management systems in digitalized buildings.

A Gap in Time: The Challenge of Processing Heterogeneous IoT Data in Digitalized Buildings

TL;DR

This study investigates the diverse dimensions of IoT data heterogeneity in both intrabuilding and interbuilding contexts, examining their implications for predictive modeling and advocates collaborative efforts to establish high-quality public datasets.

Abstract

The increasing demand for sustainable energy solutions has driven the integration of digitalized buildings into the power grid, leveraging Internet-of-Things (IoT) technologies to enhance energy efficiency and operational performance. Despite their potential, effectively utilizing IoT point data within deep-learning frameworks presents significant challenges, primarily due to its inherent heterogeneity. This study investigates the diverse dimensions of IoT data heterogeneity in both intra-building and inter-building contexts, examining their implications for predictive modeling. A benchmarking analysis of state-of-the-art time series models highlights their performance on this complex dataset. The results emphasize the critical need for multi-modal data integration, domain-informed modeling, and automated data engineering pipelines. Additionally, the study advocates for collaborative efforts to establish high-quality public datasets, which are essential for advancing intelligent and sustainable energy management systems in digitalized buildings.
Paper Structure (27 sections, 4 figures, 1 table)

This paper contains 27 sections, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Building IoT Data
  3. Knowledge Graph
  4. Data Stream
  5. Public Accessible Dataset
  6. Challenges for Intra-Building Scenario
  7. Irregular Granularity
  8. Long Tailed Distribution in IoT Device Category
  9. Heterogeneous Distribution in IoT-generated Data Streams
  10. Complex Dependencies in IoT-generated Data Streams
  11. Challenges in Inter-Building Scenario
  12. Spatial and Structural Differences
  13. Inconsistent Data Representation
  14. Case Study: Benchmarking on Building IoT Forecasting
  15. Problem Formulation
  16. ...and 12 more sections

Figures (4)

  • Figure 1: The schematic graphic about data communication in a digitalized building. The activities in a digitalized infrastructure are typically monitored by IoT sensors. Meter readings are uploaded to the Building Management System and then stored in the cloud-based server. This is a bidirectional conversation where the decisions made based on the data analytics results can intently impact the building operation.
  • Figure 2: The statistics about the count number for each unique IoT category based on Brick ontology definition. All IoT categories are grouped into Alarm, Parameter, Limit, Status, Setpoint, Point, Command, and Sensor based on the device functionalities. The assignment of each unique IoT category to the corresponding group are represented by multiple colors as legend.
  • Figure 3: Distribution analysis for data stream segments from Current Sensor, Usage Sensor and Outside Air Humidity Sensor. Each row indicates the original data stream, Kernel Density Estimator fittings, Autocorrelation Function measuring, and Fourier Transformation from top to bottom.
  • Figure 4: Visualization of hierarchical dependencies in the BTS-B building model. The inner circle represents the root of the hierarchy, with each successive outer layer iteratively mapping the top-down path from the root Building ontology through Location components, distributed Equipment, and finally, monitored Point sensors. The left side illustrates the overall hierarchical structure of the building, while the right side zooms in on a specific component Floor. This model structure is customized and varied by the unique design and IoT distribution in each registered building.