Gazprom Neft’s anti-COVID-19 programme

Read more

Finding Digital Oil

Gazprom Neft launched over 150 new digital initiatives and produced the first “digital oil” by applying the machine learning algorithm. By 2025, the company expects the digital transformation to reduce the time and cost of geological exploration by 30%.

Andrey Belevtsev, CDO, Gazprom Neft

In the interview for BGS Talks, Andrey Belevtsev discusses the importance of reducing cognitive biases in the decision-making process, Digital Oil Project, Gazprom Neft digital investments strategy, partnerships with startups, why the support of top management for the digital projects is so vital for the corporate culture, and a variety of other topics.

— In 2018 companies in Oil and Gas invested more than 1,5 trillion dollars in digital transformation and only 30% got an outcome. How do you measure the success of your initiatives?

— For the early R&D stage we created the internal strategy that we called Digital Technology Vision. This helps us to structure our early-stage R&D. We are now able to do a lot of tiny projects, 150 projects a year, with the budget allocated towards R&D for digital technologies of early maturity. So this is the way for us to try but not to spend too much before the technology is ready for further implementation. If you want to launch a digital initiative, you also have the easy investment process but the budget and the time for the initial stages are limited.

— Big Data, AI, cognitive computing, everything is made for reducing time for decision-making. How does Gazprom Neft use these technologies?

— I don’t think that any company that is not already making significant decisions based on data can survive on the market. But what we are targeting is that we want to make these decisions faster and we want to step-by-step remove possible cognitive biases out of the decision-making process.

One of our goals is to significantly reduce our exploration time and we see that we already have models that are doing the work that previously was done by geologists in 3-4 months, we already have a set of models that can do the same work in minutes.

— Are you talking about the Digital Oil Project that recently was completed by Gazprom Neft?

— We were applying the machine learning algorithm with our partner to process the geological information and it’s still in a working mode, we’re not saying that it’s fully completed but the results that we currently have are positive. We found some of the intervals that were missed in the previous times and we have the algorithm that predicted that it exists and we did a production, drilling, and we actually find oil in places where it was not found before.

Watch the full interview at BGS Talks YouTube channel.