The energy transition and digitalization are common buzzwords in the oil and gas industry. According to the head of an industrial data software firm, the former will rely on the latter.
“Digitalization will be the backbone of the energy transition,” said John Markus Lervik, CEO and co-founder of Cognite. “Instant access to trustworthy contextualized data for better business decisions will transform the industry. Those who invest in their digital agendas and focus on scalability and democratized user enablement will continue to lead the way and increase sustainability, safety, effectiveness, and profitability in the process.”
The oil and gas industry’s tribulations from the past year have also given it more clarity regarding how to embrace digitalization, said Lervik. Read on for his insights on how the industry’s focus on digitalization has changed – including his answer to the question in the title about artificial intelligence (AI) – and related trends in the digital evolution of oil and gas operations.
Looking back to early 2020 versus now, how has the oil and gas industry’s interest level in remote operations changed?
The pandemic has really applied a stress test on digitalization investments and very quickly revealed which investments were valuable and future-proof. A lot of industrial companies we work with experienced challenges of their teams having limited access to industrial sites, having to cut budgets while increasing ROI (return on investment) from digital investments, and of course turbulence related to uncertainty in the market.
However, we’ve seen the frontrunners invest in digital strategies that focus on scalability and end-user enablement – whether it’s for the engineers, the data scientists, or the domain experts in the field.
The ability to quickly scale solutions across industrial assets is a competitive advantage.
One of Cognite’s clients, which operates 30 oil platforms with more than 300 wells, lacked a unified overview of maintenance activities and the ability to communicate them effectively to workers. In a matter of a couple of months, we helped them develop and deploy a maintenance planner application powered by contextualized data from our industrial data platform, Cognite Data Fusion®, to optimize efficiency and reduce waste by enabling efficient scoping, planning and execution of maintenance work.
The application leverages contextualized data from different source systems to display a list of work orders prioritized by risk factors and maps them to a 3D model of the platform. This solution has been readily scaled from one platform to 21 platforms, helping the client to reduce planned shutdowns by close to 30%, boosting production by approximately 700,000 barrels a year. These efficiency gains are valued at $38 million a year.
Another focus area is investing in technology, which enables end-users to make better data-driven decisions by exposing them to the contextualized data in the format they understand. Our belief at Cognite is that technology is there to abstract complexity and democratize access to data. An example of a company that has really embraced this is OMV (OTCMKTS: OMVKY) as they are focusing on live operational digital twins to power remote operations.
Which segments of the industry have been most active in integrating AI, and what have they observed by doing so?
First, it is key to understand that without access to contextualized, governed data, AI in the industry is at best just a nice theory. As explained later, industrial AI indeed has a lot of potential, but one needs to make sure one has a robust data architecture, which contextualizes data from the common siloed data systems, and makes this data available with sufficient quality and performance.
As for practical uses of AI, there are two areas I can highlight that have had the most impactful results, and one emerging area that I believe will deliver a lot of impact going forward.
The first one is predictive maintenance. A number of our clients have been leveraging Cognite Data Fusion® to enable solutions for predictive maintenance by contextualizing vast amounts of historical and real-time data, analyzing it, and applying AI to predict where the next downtime is most likely to happen and enabling the companies to plan their maintenance much more proactively and actually allocate time to attend to the most critical tasks first.
The second one is production optimization. There the power lies in what we call hybrid AI which is really merging physics-based modeling with AI to forecast and optimize production. OMV is an example of one of the leaders in the field. They use Cognite’s application called BestDay. BestDay uses data-driven AI modeling to calculate what its name suggests: the best possible production day. BestDay evaluates a field’s production history, custom boundary conditions, and production criteria to produce a maximum capacity algorithm that updates daily. This creates real-time organizational visibility on all aspects of production, enabling experts to reduce unplanned deferment events and increase overall production throughput.
One of the emerging AI applications with the most impact is within the robotics field where you can incorporate data (visual and sound) – for instance, collected by robots or drones – into continuously enriched digital twins and then use this to analyze asset performance. Then you can start to automate some of these missions using Cognite Data Fusion®. You may have seen that we recently deployed Spot the robot dog on his first autonomous mission offshore aboard the Skarv in the North Sea with Aker BP (FRA: ARC).
Which types of operations are best suited for robots and other AI technologies?
Everything that happens in the industry can be optimized by data. The next big space with a lot of potential is using data to make operations more sustainable by attacking areas like fuel consumption, chemical use, and others. Here data can be leveraged to directly decrease the environmental footprint of the operations while at the same time maintaining and often increasing profitability.
Cognite is currently working with the Center for 4th Industrial Revolution for the Oceans, which is established by Aker and the World Economic Forum, together with industrial partners such as Microsoft (NASDAQ: MSFT). Here, we develop a next-generation Discharge and Emissions Tracker, where the objective is to build a digital application that will optimize, track, and eventually develop better and more efficient practices for oil and gas chemical consumption and discharge. The tracker is a joint effort between the Center and project partners, including Cognite, and the E&P company Aker BP.
Aker BP provides the tracker with real-time data and operational expertise from its offshore facilities, while Cognite is responsible for the software solutions. The tracker will deliver value in three ways:
- First, it provides process engineers with a digital tool to monitor and minimize emissions to air and release of chemicals to sea, ensuring optimal efficiency to minimize the environmental footprint.
- Second, it enables operational optimization across assets.
- And, third, it increases transparency to authorities and other stakeholders through seamless reporting.
How is AI influencing the human/workforce element in oil and gas operations? Where have deployments gone smoothly, and where have there been hiccups?
Industries need to create a culture that is fluent in technology and committed to sustainable transformation. Empowerment of workers to leverage new work processes, technology, and operational models to meet the new challenges head-on is key.
We have seen workers embrace AI solutions using our industrial data platform as it gives them instant actionable information to help them do their job. For example, thousands of operational sensors feed data hubs where crews working on an oil platform have millions of data points to consider while operating and maintaining the rig. Most legacy systems on those rigs silo data and information critical for operations, maintenance, investment, and efficiency decisions. Using AI through digital twins, photogrammetry, physics simulations, and robotics provides real-time solutions and democratizes access to these assets on the rig. Duties that are driven by routine or repetition are freed up so workers can focus attention and creativity elsewhere.
Are there any common misconceptions about AI in oil and gas that you’d like to correct?
AI is not going to replace people, but rather empower people to both be more effective and also – importantly – free time for more creative, higher-value activities.
To assist with this transition, we at Cognite have developed an industrial digital academy, targeted at both blue-collar and white-collar workers. This was initially available in Norway and is now being rolled out internationally – including in the Middle East through our collaboration with Saudi Aramco.
Looking onward, what are some facets of AI that we should increasingly see in oil and gas? How can individual firms within the industry integrate these changes?
I believe we will continue to see democratization of data use, meaning that vast quantities of data – live and historical – will be available for decision-making for everyone in the industry. In the same way that digitalization has abstracted complexity from our private lives – it will do that in the industry, too. We are able to connect our smartphone to the light switch at home not because we know how to code, but because the technology is abstracting the complexity and enabling us to do it through a simple visual user interface. The same will happen in the industry.
We will also see the true scale of AI in the industry that is going to be enabled by contextualized data and workflow orchestration and empower new ways of working.