Implementing Data Science and Intelligent Automation at Equinor

Ahead of the Intelligent Automation in Oil and Gas Summit 2020 in Huston, we sat down with Fakhri Landolsi, Head of Data Science North America at Equinor. Fakhri has lead transformational end-to-end data science projects in multiple industries; including oil and gas, both service and majors, but also automotive and robotics.

Fakhri has worked on several cutting-edge applications: inverse problems, purpose-built robotics, data-driven Diagnostics and Prognostics and Autonomous Vehicles. His focus has always been similar; creating value and leveraging all available data, moving from insights to actions. He is currently the head of Data Science North America in Equinor with exactly that same focus.


You can click here to download the whole, in-depth interview. Here is a snippet from our interview, to give you an insight into some of Fakhri’s industry insights:


Can you tell us about Equinor’s COO Digital Centre of Excellence?

The Digital Centre of Excellence was established in mid-2017. It serves as an innovation hub, driving the digital agenda of the company. Centralizing digital efforts helps with exploring emerging technologies, such as Advanced Analytics and AI.

The most important thing is to connect these technologies with real business challenges and needs. The DCoE is driving digital opportunities via three technological enablers: robotics, process digitalization, and data science with a focus on expanding the scale of our digital operations.

How are you employing data science for intelligent automation at Equinor?

We are approaching data science problems in Equinor from a platform perspective. We solve different problem sets with a common data science “mold”. As a rewe call it epsilon scaling. This includes our approach to predictive maintenance, operational planning, and in fact anything and everything. It is really a focus on scale and speed to build real business impact.

At this year’s IQPCIntelligent Automation in Oil & Gas Summit, you’re going to talk about how to gain buy-in and fully leverage the power of analytics. Can you elaborate on that?

This is the real challenge that the whole industry is facing. It’s good to develop data science solutions, but really, the value comes from moving from insights to actions. You have to ensure that the business uses these tools in their operations to realize an impact. That might be easy for machines; it’s not that easy for people and processes.

It’s really about a cultural change and emotional intelligence. In a sense, you should start small but think big. We build trust by creating quick MVPs and POVs. As a result, operators are not afraid of making decisions based on those tools. It’s a buy-in with the mindset of collaboration and an understanding of the “as is” and the “to be.”

Can you give us an example of one of the small quick-win projects that you just spoke about?

An example is a product that uses Machine Learning to forecast production of unconventional wells. Classically, the problem was approached in a linear way. If I’m planning to forecast the production of the next 10,000 wells, then I will need the same number of reservoir engineers I needed for the first 10,000 wells.

ML allows us to automate the process while incorporating subject matter expertise. This saves time for engineers to focus more on what matters and unlock the potential to run multiple “what-if” scenarios for future wells. Building trust on the results of the existing wells ML forecast is the first step. Which should be, really, a quick win, and then you keep scaling, and you build on top of that.

You can read the whole interview with Fakhri here. Click below to find out more about the Intelligent Automation in Oil and Gas Summit 2020 in Huston, February 24-25, 2020 at the Norris Conference Center.