Exploratory Data Analysis on Open Heterogeneous Building Occupancy Datasets

The development of smart buildings and advanced technologies such as Internet of Things (IoT) systems, smart sensors and data analysis methods play a crucial role in achieving a higher level of societal energy efficiency. The optimization of energy consumption by evaluating the number of occupants in buildings and assuring personalized comfort conditions requires the integration of these technologies, promoting innovation in building automation systems, through energy management and control systems. The paper focuses on a detailed analysis of open heterogenous building occupancy datasets from the perspective of a robust data science pipeline. We introduce our methodological contribution, mainly in the form of data pre-processing and standardisation, for the analysis of diverse data types utilized for occupancy estimation in buildings. The proposed algorithm is based on an examination of timely open datasets containing building sensor data, offering a well-informed reference for occupancy estimation and forecasting algorithms.