Reimagine your geospatial analytics workflows for subsurface data in the era of generative AI

Overview

The combination of AWS generative AI, S&P Global Enterprise Data Management (EDM) for Energy, and Esri ArcGIS tools orchestrates a harmonious symphony in subsurface exploration workflows within the energy sector. Generative AI, with its adept artificial intelligence (AI) capabilities, composes intricate melodies by exposing complex geological datasets to non-geotechnical audiences. EDM for Energy serves as the conductor, facilitating a seamless flow of data, harmonizing the organization, quality improvement, and accessibility of critical information through Energy Data Insights (EDI) on AWS for the OSDU Data Platform. The integration with Esri ArcGIS tools, made possible for the OSDU Data Platform Geospatial Consumption Zone (GCZ), transforms this subsurface data into a visual masterpiece of spatial understanding, where geological features such as wells and well formations become key notes on a map. This collaborative approach not only optimizes subsurface workflows but also facilitates interdisciplinary collaboration, allowing geoscientists, engineers, and decision-makers to collectively compose a more nuanced exploration narrative. The resulting subsurface symphony represents a transformative crescendo in the energy industry’s digital evolution, unlocking new dimensions, breaking down silos, and creating business synchronization in decision-making.

Subsurface data management

Data management remains a key area of focus in the energy industry workflows, where the growing amount of subsurface, regulatory, production, and operations data continue to challenge companies. However, as the digitalization in the energy industry progresses and technology evolves, common workflows can be re-imagined. In this case study, we illustrate how the enterprise data can be unlocked and analyzed not only by technical audiences, but also by the audiences that might not have a deep technical background.

The growing amount of subsurface data presents significant challenges in data management, storage, integration, quality assurance, processing, interpretation, visualization, and collaboration. Massive datasets from well logs and core samples demand robust data management systems and substantial storage capacity. Integrating heterogeneous data formats and sources needs standardization and robust tools to maintain consistency and compatibility. Addressing data quality issues, uncertainties, and noise is crucial for accurate business outcomes. As data volumes increase, processing and interpretation become computationally intensive, necessitating advanced algorithms, high-performance computing resources, and efficient workflows. Effectively visualizing and communicating complex subsurface information to diverse stakeholders needs intuitive techniques. Data sharing and collaboration among organizations face hurdles because of proprietary concerns and the need for secure services. Moreover, the growing demand for advanced data analytics and generative AI and machine learning (ML) techniques to extract insights and support decision-making processes needs specialized expertise and computational resources. Addressing these multifaceted challenges necessitates interdisciplinary efforts involving geoscientists, data scientists, computer scientists, and domain experts.

OSDU Data Platform and EDI

Recognizing the need for a unified approach, the energy industry organizations embarked on a quest to establish a single source of truth (SSOT) repository, a centralized haven for their subsurface data. This endeavor promises to bring order to the chaos, providing data consistency, integrity, and reliability across the entire enterprise. The OSDU Data Platform, based on the energy industry standards, allows the data to be quickly accessible by a variety of applications. The EDI for the OSDU Data Platform on AWS solution brings fully managed product experience for production-grade workloads that removes the complexities of managing and deploying the OSDU Data Platform yourself. The conceptual architecture of the EDI on AWS for the OSDU Platform is shown in the following figure.

Figure 1. Conceptual EDI architecture

EDI on AWS offers several benefits for organizations in the energy industry. AWS provides a robust and scalable cloud infrastructure that can handle the massive amounts of data generated by oil and gas operations. The OSDU Data Platform can use the power of AWS to streamline data management, enhance collaboration, and drive exploration. One of the key advantages of EDI on AWS is scaling computing resources on-demand. As data volumes and workloads fluctuate, organizations can quickly scale up or down their AWS resources, making sure they only pay for what they need and avoiding over-provisioning or under-provisioning. Furthermore, AWS offers a wide range of managed services, such as Amazon S3 for object storage. EDI on AWS supports the Amazon S3 intelligent-tiering, designed to deliver cost savings by storing data in the most cost-effective tier while making sure that frequently accessed data remains readily available.

Mastering subsurface data

Before the SSOT, such as OSDU Data Platform, can be constructed, reconciling data that is available from various data sources is paramount. Using configurable, off-the-shelf plug-ins for common subsurface applications, we can blend, master, compare, prioritize, govern, move, expose, and improve the quality of data with support for common geo data types, such as well headers, directional surveys, formation tops, logs, geochemical analysis, and more. S&P Global EDM for Energy can help achieve this. The overview of mastering the data is shown in the following figure.

Figure 2. EDM for Energy solution

S&P Global EDM advanced data management tools enable users to curate, enrich, and harmonize data, facilitating consistency and accuracy across different datasets. This streamlined data management approach minimizes errors, reduces redundancies, and enhances data reliability. The EDM for Energy workflow also supports data exploration and analysis capabilities. Users can use powerful visualization tools, interactive dashboards, and advanced analytics to gain insights from their data. When the data is mastered, it is ready to be ingested and cataloged in EDI on AWS for broader use cases and consumption, such as spatial analytics and generative AI.

Geospatial analytics

Showing data on a map is a fundamental requirement for effective data management in the oil and gas industry. This is because the industry deals with a vast amount of geospatial data, such as seismic surveys, well locations, pipeline routes, and production facilities. Visualizing this data on a map provides spatial context and helps in understanding the relationships between different data points and their geographical locations.

Esri’s ArcGIS Experience Builder is a powerful tool that can greatly enhance how oil and gas companies use maps and geospatial data for their operations. With ArcGIS Experience Builder, companies can create interactive and customized web applications that combine their geospatial data with other relevant information. These applications can be tailored to specific use cases, such as exploration planning, asset management, environmental compliance, risk assessment, and stakeholder communication.

Connecting the robust Esri ArcGIS capabilities is possible through the OSDU Data Platform GCZ, which comes with EDI on AWS. The GCZ creates a synchronized geospatial index of your OSDU data, such as wells, logs, markers/tops, seismic, documents, horizons, and more delivered in ready-to-use map services. Then, these map services can be used with low code/no code app solutions, where one can have a view of their SSOT data on a map and query it spatially, or even augment it with more non-OSDU map layers.

Unlocking the data for broad use cases

We can dive into the wells data example in more detail, combining the preceding approach and adding a generative AI component that makes the data searchable and queryable for broad audiences. To accomplish this, we use Amazon Q, a generative AI-powered assistant. As mentioned previously, the wells data can be seamlessly mapped with the Esri ArcGIS capabilities through the OSDU Data Platform GCZ, which comes with EDI on AWS. However, the underlying formation tops and other correlated datasets might not be readily queryable for non-technical users. This is where new generative AI capabilities can significantly help. The data can be indexed ahead of time and the queries can be done in a natural language, without the need to have the knowledge of JavaScript Object Notation (JSON) or OSDU Application Program Interfaces (APIs). When you need to interrogate the data, it could be as direct as writing a question in the chat window. An example of this is using Amazon Q to find well formations that are similar across multiple wells in a certain Texas county. The example of the user interface for this scenario is shown in Figure 4. This rather direct query is just the beginning of how Amazon Q is adding a verbal interface to overall data exploration and discovery. Amazon Q is used for the specific business need while protecting the privacy of your data.

Instead of having a geoscientist or an engineer to sift through the data through standard queries, table joins, or Python scripts, having your corporate trusted data in one place opens up an opportunity to connect the latest cloud technologies such as Amazon Q. It can be quickly integrated in the overall workflow to enable generative AI capabilities. The architecture of the overall solution is shown in the following figure.

Figure 3: S&P Global EDM, Esri ArcGIS, EDI, and Amazon Q solution architecture

The solution shown in the preceding figure is broadly applicable to many subsurface data workflows, but it can especially be of interest to non-technical and interdisciplinary audiences. It allows you to quickly interrogate your enterprise data in the area of interest both on the map and by analyzing related datasets at the same time.

Figure 4: Esri ArcGIS Experience Builder Application with Amazon Q generative AI embedded assistant for subsurface data

Summary and conclusions

The use case and the preceding workflow demonstrate how the subsurface data can be made accessible to a broad audience, independent of their technical proficiency or familiarity with the specific region of the data itself. The whole workflow is designed with almost no coding, chaining together S&P Global EDM for Energy, EDI, Esri ArcGIS, and Amazon Q with direct no-code/low code interfaces. When the initial solution is set up, it is also readily extensible to more scenarios and data types. Directional surveys, well logs, and other data types can be added in a similar fashion by using S&P Global EDM for Energy, allowing mastering and blending these data types across multiple sources before uploading them to the enterprise data catalog. If this catalog is based on the energy industry standards, such as EDI on AWS for OSDU Data Platform, then the rest of the integrations are direct, using the standard APIs and data definitions. As the solution is powered by the secure and reliable AWS Cloud infrastructure, it is scalable and can accommodate datasets of virtually any size.

Managing subsurface data is an involved process in any organization, but the tools and processes are rapidly evolving to overcome data management challenges in the energy industry despite increasing complexity and the ever-growing amount of data. Mastering the disparate data sources using EDM for Energy builds trust in your enterprise datasets. Then, the OSDU Data Platform seamlessly integrates these datasets with a variety of applications and tools, and OSDU GCZ further unlocks the data for rich geospatial analytics applications and workflows. With this data, you can build maps and applications with little to no coding. These new technologies and seamless integration options free up time for you to focus on gaining valuable insights.