RSE IFC CIM Work - statnett/Talk2PowerSystem GitHub Wiki

  • CIGRE Paris Session 2026 D2 INFORMATION SYSTEMS, TELECOMUNICATIONS AND CYBERSECURITY - Full Papers PS1 - IT/OT Solutions to improve the Efficiency and Resilience of Electric Power Systems ID: 10398 Zotero [eneabiondaDigitalTwinAsset2024] Enea Bionda, Gabriele Paludetto, Francesca Soldan. RSE, Italy
  • Digital twin for asset management of electric power systems based on IEC CIM and BIM integration Price: 30 EUR Il contributo di RSE all’edizione CIGRE 2024 Francesca Soldan presented the poster “Digital twin for asset management of electric power systems based on IEC CIM and BIM integration” by RSE (G. Paludetto and E. Bionda). The paper describes the use of ontologies and knowledge graphs for the integration of different information models useful for the representation of electrical network components under different perspectives. In particular, the combination of the IEC Common Information Model and Industry Foundation Classes (relating to the Building Information Modelling methodology) is presented to overcome interoperability problems in data management throughout the life cycle of network assets. A slide shows "GenAI opportunities in electric power industry"
  • Francesca Soldan posted on LinkedIn Delighted to have participated in CIGRE Paris 2024! It was an incredible opportunity to connect with power system experts from around the world and share insights from our latest research. I had the privilege of presenting our paper, “Digital Twin for Asset Management of Electric Power Systems based on IEC CIM and BIM Integration”, co-authored with my colleagues Enea Bionda and Gabriele Paludetto. Our work is dedicated to integrating the IEC Common Information Model (CIM) with Building Information Modeling (BIM) to improve data management throughout the entire lifecycle of grid assets. By breaking down informational silos, we aim to foster semantic interoperability and enhance asset management efficiency.

Slide 10: particularly interested in harmonization of CIM with IFC and hey use GraphDB and ONTOP virtualization, and timeseries!

Zotero: [paludettoRealCaseStudy2021] Real Case Study of Electrical Grid Conversion from Non-standard Data Models into IEC CIM

Chimera

Chimera (GitHub).

A software suite that aims at better connecting the big data world with semantic technologies. It provides two components for enabling Knowledge Graphs empowered analytics scalable to big data technologies:

  • OntopSpark: an extension of the Ontop Ontology Based Data Access (OBDA) system, which uses Apache Spark as a query processing engine for accessing the data stored in data lakes. The integration of a distributed data processing engine such as Apache Spark allows exploiting the Ontop data integration capabilities at its maximum potential, as it brings all the advantages of velocity and parallel computation typical of a distributed system to the task of solving a SPARQL query.
  • PySPARQL: a library that allows the users to query a SPARQL endpoint using a python notebook, process the response with Apache Spark, and eventually store the Spark DataFrame / GraphFrame into the data lake.

OntopSpark

An extension for enabling Ontop to perform OBDA by querying relational data physically stored in Spark tables. OntopSpark uses Ontop's Virtual Knowledge Graph (VKG) mechanism, which allows creating RDF representations of relational data without allocating additional space. OntopSpark needs three files to respond to SPARQL queries by building the corresponding VKGs: 1. a DB-descriptive ontology file (usually .owl or .ttl) containing the ontological concepts in OWL2QL profile needed by the Ontop reasoner, the semantic of the relational data stored in the Spark tables, 2. a configuration file for correctly instantiating a JDBC connection to the database, and 3. a mapping file for the RDF-to-SQL translation of the VKGs. For more details on how to configure OntopSpark, see the configuration section. OntopSpark is available in two different packages, namely OntopSpark-Protégé and OntopSpark-CLI. The first one is an extension that allows building ontologies and mappings using the graphical interface of Protégé, while the second exposes a SPARQL endpoint (web GUI or CLI) used for industrial deployment. We use the OntopSpark-CLI package for building a docker image available on DockerHub.

SPARQL queries using notebooks and Apache Spark

The PySPARQL python module allows the users to query a SPARQL endpoint using a python notebook and to process the response inside Apache Spark. PySPARQL leverages pyspark to manage Spark DataFrames, and uses well known libraries such as SPARQLWrapper and rdflib to handle the communication with a SPARQL endpoint and manage the result. PySPARQL is tested with multiple Spark versions and is available on PyPi.

The library retrieves the results and materializes them inside the configured Spark Session. Users shall specify the endpoint configuration at initialization time and change it during the program execution. In particular, the output type directly depends on the SPARQL query type.

SELECT queries return Spark DataFrames in which the columns directly correspond to the variables declared in the SPARQL query. However, PySPARQL does not convert the value types. The users can then process the DataFrame inside Spark and, if necessary, save the DataFrame as a Spark table.

CONSTRUCT queries return either a DataFrame or a GraphFrame depending on what the user chooses to materialize. In both cases the data resemble the constructed graph.

Papers

Digital twin for asset management of electric power systems based on IEC CIM and BIM integration

Enea BIONDA, Francesca SOLDAN, Gabriele PALUDETTO, CIGRE 2024 Paris

Problem - efficient asset management To oversee the complete asset lifecycle, from initial design to decommissioning, while preventing information silos

Solution - informative digital twins (CIM+IFC)

  • CIM facilitates data exchange within utilities, but lacks specific classes for planning and managing grid assets
  • IFC standard and BIM methodology are specific for asset management but are not specialized for the description of electrical grid components

A DT is defined as the virtual representation of a device or system that is continuously updated with real-time data throughout its entire lifecycle. It utilizes simulation and artificial intelligence techniques to contribute to and support various decision-making processes, such as operation or preventive maintenance of the device/system. DT distinguishes itself from a simulator by integrating, in addition to a mathematical model of the reality to be emulated and a simulation environment, a mechanism for updating the parameters of the mathematical model, thus keeping it constantly aligned with the actual state of the reality being emulated throughout its lifecycle.

Event-Based Resource Management Framework for Distributed Decision-Making in Decentralized Virtual Power Plants

J. Zhang; B.C. Seet; T.L. Tek, Energies, vol. 9(8), no. 595, 2016.