Open Source for Energy Transition via Data Management - Eyes on a Major Prize

### Overview In this talk, Ati, Digital Commercial Lead and Open Source Lead at Shell India, details how digital technology, open-source software, and unified data management are essential to accelerating global energy transition. Drawing on findings from an independent MIT Technology Review Insights report surveying 350 C-suite executives across eight sectors, digitalization is highlighted as one of the fastest and most cost-effective levers for decarbonization. Shell advocates for moving beyond traditional siloed industry practices toward boundaryless, open ecosystems—collaborating closely with organizations like the Linux Foundation Energy (LF Energy) and industry partners to standardize data and build shared foundational infrastructure.

### Key Topics - **Digitalization and Energy Transition**: Decarbonization across complex value chains relies heavily on AI and data optimization to manage variable renewable sources, integrated energy systems, and decentralized low-carbon networks. - **Data as the Common Equalizer**: Diverse assets—such as solar farms, refineries, hydrogen plants, and electric vehicle infrastructures—all generate time-series data. Unified data governance, data mastership, and standardization are required to enable interoperability. - **Open Source Collaboration**: Transitioning from proprietary protectionism to open-source initiatives to solve shared decarbonization challenges, exemplified by LF Energy participation and internal initiatives like shell.ai. - **Open AI Energy Initiative (OAEI)**: A collaborative marketplace co-founded in 2021 by Shell, C3.ai, Baker Hughes, and Microsoft to integrate tools (such as predictive maintenance and Kongsberg digital twins) and share foundational technologies. - **The Real-Time Data Ingestion Platform (RTDIP)**: A core time-series data ingestion tool developed by Shell and contributed as an open-source project under LF Energy.

### Technical Details - **RTDIP Architecture**: Built for processing high-velocity time-series sensor data in cloud environments, streaming data via Event Hub or Apache Kafka directly into a Delta Lakehouse. - **Developer and User Interfaces**: Features an RTDIP Python SDK and REST APIs designed for deploying queries, executing workloads, and managing security, making ingestion accessible across technical domains. - **Operating Scale**: Deployed across Shell assets globally, ingesting data from over 3 million physical sensors and scaling to more than 4 trillion rows of data. - **Integrated System Applications**: - *Electrified Steam Crackers*: Electrifying a single cracker can eliminate emissions equivalent to 350,000 European cars, requiring complex real-time grid integration. - *Multi-Scale Modeling*: Combining physics-based models with AI to integrate intermittent solar and wind generation with industrial hydrogen electrolyzers (such as Shell's Rhineland refinery in Germany) and cement plants. - *Thermal & Fuel Innovations*: Leveraging RTDIP and time-series simulations for Shell E-Fluids (providing 30% improved battery cooling in partnership with Grayson), synthetic kerosene aviation fuel produced from CO2 and water, and liquid hydrogen storage and transport selected under a U.S. Department of Energy (DOE) project.

### Takeaways To achieve net-zero climate goals and effectively track Scope 1, 2, and 3 emissions, industries cannot rely on thousands of competing, incompatible platforms. Converging on common open-source standards and data platforms like RTDIP enables seamless interoperability across both energy producers and hard-to-abate sectors like cement and steel.