For Shell, AI and data are as important as oil

At Shell, there are many reasons to use AI and data to transform their business.

From increased energy demands and disconnected environments to increased pressure to fight climate changeThe oil and gas industry is at a crossroads. Energy companies like Shell can stick to the status quo or embrace the idea of ​​a low-carbon energy future.

The transition towards a more distributed, diverse and decentralized energy system means optimizing end-to-end processes and maintaining them at scale. That means solutions that can be deployed globally at lightning speed are crucial. And it meant that Shell had to become an AI-based technology company.

Accelerate digital transformation

For example, last November, Cover establish Open AI Energy Initiative (OAI) along with Baker Hughes, Microsoft and enterprise AI company C3 AI to help accelerate the energy industry’s digital transformation.

According to Dan Jeavons, vice president of computer science and digital innovation at Shell, OAI offers industry leaders the opportunity to collaborate openly, fairly and transparently. It allows them to create interoperable standards between AI applications and accelerate digital technology adoption and achieve net zero emissions in the future.

“We have committed to being unreal by 2050 or earlier and achieving 50% reductions in emissions of scope one and two by 2030,” he said.

While digital technology may not be a silver bullet, it is one of the core levers Shell is using to accelerate the energy transition. Jeavons adds, “While we will need to transform a lot of hardware to transform the energy sector, we can also take advantage of the data we have today and use that data to transform system.”

AI plays an important role in Shell’s business strategy

Shell has implemented a number of AI initiatives over the years, including implementing reinforcement learning in its exploration and drilling program; deploying AI at public electric car charging stations; and install computer vision support cameras at service stations.

The company also recently launched the Shell.ai Residency Program, which allows data scientists and AI engineers to gain experience working on a variety of AI projects across all of Shell’s businesses.

Currently, Shell is deploying north of 100 Applications AI into production every year. They have also developed a central community of more than 350 AI professionals who are designing AI solutions using the massive data source available across multiple businesses within Shell.

AI helps Shell predictive maintenance

“Reliability and safety are absolutely fundamental,” says Jeavons. “Being able to recognize when things are going wrong and proactively intervene is our top priority.”

AI has enabled Shell to use predictive surveillance to enhance the surveillance techniques they already have.

To give that perspective, Jeavons claims they have more than 10,000 devices currently being monitored by AI – from valves and compressors to dry air seals, gauges and pumps, while the AI ​​also makes predictions about incidents may occur. To monitor all of those devices, 3 million sensors collect 20 billion lines of data per week, while nearly 11,000 machine learning models allow the system to make more than 15 million predictions per day.

Historically, Shell relied on physics-based models to make these predictions. Before the C3 AI-powered predictive maintenance program, the company would usually replace parts after a certain period of time. This approach means that parts are often replaced while they are still in good condition. An alternative strategy is to wait until something fails. With equipment breakdown, the property needs temporary closure for repair, affecting production.

AI-driven predictive maintenance has enabled the company to reduce equipment and maintenance costs by using resources more efficiently, reducing production disruptions, and avoiding unplanned downtime.

Tom Siebel, CEO of C3 AIexplains that there are a lot of infrastructure and coordination issues around AI.

“It’s not that hard to build machine learning models,” he said. “The hard part is getting two million machine learning models into production, into one application.”

However, with active engineering monitoring, Shell data scientists can analyze thousands of data points simultaneously and enable engineers and others to derive insights from the data. whether that.

“Our team uses that data to understand what normal behavior on the basis of our assets looks like under specific circumstances, including equipment like compressors, valves, and pumps.” Jeavons said. “We then create projections of what we think is normal in the coming periods. From that forecast, we can determine when normal conditions no longer occur and then link that back to historical events.”

Next is AI to optimize for Shell

Now, Shell has commercialized its AI predictive maintenance applications built with C3 AI software. Going forward, Jeavons said the company is currently focusing on laser optimization.

“This means we can identify ways to produce more efficiently, produce more output at the same cost, and more importantly, we can also look at the CO2 of these processes. and start optimizing accordingly,” says Jeavons.

In the near future, Shell is also exploring how AI can be leveraged to monitor carbon capture, storage installations and methane levels, he added.

“These joint ventures involve making our existing business more efficient and effective, but also play an important role in our energy transition strategy,” he said.