JüKa BuilData: Homogenized Heterogeneous Data from Building Operation

"Building operations are contributing almost 30% of global CO2 emissions, highlighting the need for intelligent and adaptive operation strategies that can reduce and control these emissions. However, effectively training and developing AI\/ML models to address these challenges requires large amounts of fully labeled real-world building data, which is currently scarce and difficult to obtain. To address this problem, J\u00fcKa BuilData provides a collection of high-fidelity, time-series designed to benchmarkadvanced control strategies for Building Energy Management Systems (BEMS). The data is collected from two real-world laboratories, under the framework of the Living Lab Energy Campus (LLEC) project, operated at Karlsruhe Institute of Technology (KIT) and Forschungszentrum J\u00fclich (FZJ), in Germany in 2025. The primary objective of the dataset is to capture the complex interplay between electrical consumption and thermal dynamics in residential buildings. The synchronized 1-minute-resolution measurements include electricity consumption, photovoltaic (PV) generation, voltage, current, indoor temperature, weather conditions, and a normalized dynamic pricing signal. ... mehrThe dataset is specifically structured to support multi-objective optimization, enabling the development and evaluation of control algorithms that simultaneously minimize operational costs and maintain occupant comfort. By providing high-resolution, real-world measurements tailored to control-oriented applications, J\u00a8uKa BuilData bridges the gap between theoretical algorithm development and practical BEMS deployment. As such, it fosters advancements in modern data-driven energy management approaches, including reinforcement learning and other machine-learning techniques, and contributes to the broader goals of energy-efficient building operation and sustainable urban development. This paper details the data acquisition methodology and file structure to facilitate its use by both the AI and power systems research communities"