Modernizing bp’s application landscape with AI

One of the oldest and largest oil and gas companies in the world, bp is in the midst of a major transition as it pivots toward becoming an integrated energy company. As part of this transition, the company is aiming for a net-zero carbon footprint by 2050. The scope of its efforts so far is demonstrated by its shift into lower-carbon businesses, power trading, and convenience stores, which represented just 3% of its investment in 2019 but 23% in 2023.

This change in business focus is accompanied by an ongoing digital transformation. A key part of this is to use AI to help improve its operations.

From COBOL to EVs: The scope of bp’s digital initiatives

As the energy industry goes through multiple mergers and acquisitions over the years, it deals with a wide spectrum of digital apps, legacy systems, and business processes across its functions. While some of these are modern, cutting-edge elements of the tech stack, they also carry the load of outdated technologies that require transformation.

In bp’s case, the multiple generations of IT hardware and software have been made even more complex by the scope and variety of the company’s operations, from oil exploration to electric vehicle (EV) charging machines to the ordinary office activities of a corporation.

Historically, AI use has been focused on machine learning in operations such as exploration and drilling in the initial phases of energy production. Now, however, generative AI (genAI) and other forms of digital innovation are helping drive efficiencies closer to the end customer.

The reality for a company of bp’s size and history is that there are different levels of technology maturity in different areas of the business. This reflects both the diversity in the technical infrastructure as well as the readiness to experiment of different operating units.

Focusing on business value before AI

Mariza Fotiou, VP for digital product management at bp, emphasizes the importance of focusing on the business value that is being sought, whether AI is involved or not. “The first thing to look at,” she says, “is the problem I’m trying to solve, and the solution I can create to generate value with the least complexity.”

In that context, she adds: “We’re not looking at AI use cases [per se]. We are looking at bp problems or customer problems that we need to solve that AI can accelerate.”

Fotiou draws on her background in product development and digital transformation—first in the finance sector and then in bp’s upstream operations—to help solve downstream challenges in the B2B space, especially in mobility and fleet operations. With such a wide variety of products in development across the company, finding the best ones for AI requires prioritization.

Fotiou and her colleagues particularly look to “parts of the business that have an intense amount of data that we’re trying to manage, because that’s where AI can drive the most value.” Engagement with leadership and upskilling for personnel help “develop the conditions for AI innovation and experimentation to take place,” she says.

Along the way, the company decides whether to build or buy a solution for each use case. To guide that decision, bp applies consistent design governance principles to find the solutions—always grounded in safety—that are most competitive, optimal in terms of cost, and likeliest to provide the company with a differentiating advantage.

“If it makes sense for us to create and maintain the IP,” Fotiou explains, “then we will go off and build something. If we are lagging and just playing catch-up, we might as well buy it. But cost is always a big part of the equation that we need to consider.”

Fotiou has found that in some cases open-source tools can help especially with cost considerations. For example, her team is leveraging open-source AI product management tools to help define thousands of product requirements as they replace their fleet management system. “We’re leveraging AI as much as we can around product development,” she says, “and it seems to be quite helpful.”

How bp Is putting AI to work

“We have found ways to use AI across all these value streams,” Fotiou says, “from helping produce energy to trading the energy, all the way to supplying and distributing the energy.”

AI has helped bp identify the best locations for placing EV charging stations for customers to use them, and it has enhanced the company’s award-winning safe2go fuel data platform, which uses computer vision to ensure that aircraft receive the correct fuel. Like many companies, bp is also using genAI to extract information from documents, summarize meetings, and so on, freeing up office workers’ time for more strategic activities. And it uses AI to automate code testing and other aspects of the digital development lifecycle.

In 2023, Infosys became bp’s main partner for end-to-end application services, helping to transform bp’s digital application landscape. Infosys, among other partners, is helping bp implement a variety of AI projects across different areas of its operations, and is working with Fotiou and her team on several legacy applications in bp’s fleet management suite. These apps have millions of lines of code; Infosys is using genAI to help analyze this code and deliver precise recommendations for refactoring, code conversion, documentation, automated test cases, and test scripts. The work accelerates feature development on the codebase and creates a robust knowledge management repository to reduce artifact creation efforts by more than 70%. This refactoring helps to set the base for digital transformation.

Solving problems for many kinds of customers

It is evident from these examples that bp is leveraging AI across many facets of its business. Whether in production management, application modernization, fleet fueling, or retail store operations, AI can help improve efficiency in delivery and help support the business transformation that drives value creation.

“This is our IT strategy: to help bp transform into an integrated energy company,” Fotiou says. “We try to focus on priorities that make sense for different areas of the company—always with a customer-centric outlook. Every area has their own priorities that we can capitalize on, and with the help of AI, become smarter, faster, more competitive in the market.”

Discover how to accelerate business value with Infosys Topaz, an AI-first offering.