RO-Crate Time Series Exporter for the Building Consumption Data of KIT Campus North

The facility management (FM) of the Karlsruhe Institute of Technology (KIT) operates an infrastructure for measuring energy consumption to invoice other organizational units within KIT for the energy consumed. For this purpose, the measuring infrastructure automatically records and stores the energy consumption of all buildings on Campus North at a resolution of 15 minutes. The recorded and stored consumption comprises different energy types, namely electricity, gas, heat, water (warm, cooling, drinking, and several kinds of wastewater), and compressed air. Since this measurement infrastructure is already in operation since 2006, the consumption data stored as time series meanwhile cover a long period of time. The covered period of time makes these energy consumption time series highly interesting for the energy research community, especially for energy researchers at KIT. However, accessing the data is challenging. While the original infrastructure was designed for single-user access and limited data throughput, it now faces multiple users and high data throughput. Moreover, since the used technology does not scale with the ever-growing data volumes, FM finally updated the data infrastructure. ... mehrHowever, despite improvements with regard to performance, the new data infrastructure brings new challenges, including data only partially moved to the new infrastructure. For this reason, retrieving time series whose time range spans data from both old and new infrastructure requires a researcher to write queries for both database systems, which in turn requires knowing the complicated logic of the used schemas of both databases. Even if a researcher successfully queried such a time series, she needs further queries to allow measurement units, measurement quantities, and scaling factors to be included in the interpretation of the data. Given both this challenging data access and the increasing interest in the data, we started to simplify the process of data querying by developing a web service with a simple REST (Fielding, 2000) interface. This interface allows researchers to query data in a unified way, without requiring any knowledge about the underlying databases and thereby lowering the hurdles of accessing the data. The interface requires only a time range, a list of buildings, and energy types as inputs and returns a ZIP file including the time series as CSV files and an RO-Crate (Soiland-Reyes et al., 2022) metadata file in JSON. The metadata file fully describes the requested energy consumption time series by using the RO-Crate data package standard with an extended, in-house developed profile for time series description. This RO-Crate metadata file enables an interpretation of the obtained data without any prior knowledge and reduces the burden on researchers to publish the data according to good scientific practice. Since a lot of research using energy consumption data benefits from including exogenous influences such as weather (Dannecker, 2015 & Haben et al., 2023), the developed web service also allows obtaining weather time series for the specified time range, which again is described in the RO-Crate metadata file. The present poster shows the steps taken to develop the web service: It starts with the analysis of the original database schemas, before it describes the agreement on the required information resulting in a shared database schema. The poster continues with the transformation of the original data into the shared schema that builds the data foundation of the service. Next, the poster presents the creation of the time series profile, the standards and vocabularies, the used technologies to develop the service, and the challenges during the development of the software. The poster concludes with an outlook on planned improvements and extensions of the developed web service.