At Wintershall Dea, we’re committed to digitization that improves our decision making and makes us more efficient, competitive and sustainable. We gained these benefits from a recent project to apply artificial intelligence (AI) to early exploration research.

Working with IBM Research in Zurich and IBM Services, we developed the Exploration Advisor Tool, or EAT, an intelligent search engine for explorationists that uses AI to analyze a vast trove of unstructured geological data and information.

The EAT solution thinks like a geologist, helping our researchers quickly discover insights from the data. This helps them to make better and faster decisions in the intense competition for gas and oil exploration contracts.

Interested in solutions for oil and gas?

The challenge of analyzing legacy data

A key challenge for explorationists is making sense of legacy data. Time is of the essence, but the data is often hidden in different folders and drives. Plus, especially in early exploration, it exists predominantly in unstructured sources such as reports, presentations, spreadsheets and operational logs.

Finding the relevant information by conventional means is time consuming. Data acquisition and analysis in early exploration is time-constrained—doing it properly can take weeks or months with significant financial and opportunity costs.

Trained to think like a geologist: Creating an AI Exploration Advisor

That’s why we engaged a multidisciplinary team from IBM to address the problem. Using natural language processing and machine learning, our specialists worked with the IBM team to train the tool to serve experts in early exploration research to reduce uncertainties in decision making.

The EAT solution organizes the data in a way that is far more intelligent than simple keyword searching. It actually understands geological concepts and what researchers are looking for, which helps deliver insights quickly. When studying a particular exploration tract, we can ask questions like “What reservoir formations are in this basin?” and get an intelligent answer.

The advisor also helps discover whether previous teams worked on similar projects and ensures that we don’t miss anything important. And by relieving researchers of mundane tasks, it frees them for higher value work.

A game changer for Wintershall Dea and the oil and gas sector

In the oil and gas sector, a key value driver is time. From acquiring licenses to starting production can take over a decade. I’m convinced that the AI-powered EAT will save us time in the crucial early phases of the value chain and help us make better predictions about expected oil and gas volumes.

In addition, the tool can be useful in other phases of the oil and gas lifecycle, such as late-stage exploration, development and production, engineering and operations—anything that depends on analyzing unstructured data. The promise is so great that we at Wintershall Dea have a new saying: “When you’re looking for data, just EAT.”