bp — 7 videos in collection
Emeka Emembolu, executive vice president, technology, at bp, speaks to Neil Shah, executive director and market strategist at Edison Group, in the third interview in Edison’s series on bp. They discuss where technology sits within bp and how it supports the company’s strategic priorities, including capital discipline, operational efficiency, downstream and trading. Emeka sets out bp’s framework for adopting AI and its key technology partnerships. He also outlines use cases across upstream, refining, retail and trading, from high-performance computing and the Palantir Foundry to the earnify loyalty platform and AI advisers for traders. Finally, he describes bp’s robotics programme at the Sangachal Terminal in Azerbaijan.
Where does technology sit within bp’s organisational structure and what role does it play in the company?
Emeka Emembolu: I report directly to our CEO and I represent the technology organisation. In my team, we cover a range of different activities: safety and security, including some intelligence; research and development for the company; digital; and procurement for the group. So it’s a range of activities.
But what these things have in common, and what our role is for bp, is really to help give bp a competitive edge – to make sure that in the things we choose to do in our different businesses, we can do them using the best technology to get the best outcomes for our company. That’s the focus, in line with the overall strategy of the company. This is what we’re here to do.
How is bp progressing strategically this year and how does technology support bp’s stated priorities?
Emeka Emembolu: For bp, a big focus for everyone is a simpler bp, a stronger bp, a more valuable bp. This is what we’re trying to do as a company. Our new CEO, in our Q2 results, laid out five key priorities, and I’d say technology has a really good role to play across the board in terms of the progress we’re making against these different areas.
On the capital side, as we said at Q2, we’re expecting our capital for the full year to come in within the $13.5–14bn range. So we’re continuing to tighten and make sure we’re driving discipline into all of the investments that we’re making. At the same time as coming in within that range, we’re strengthening and simplifying our capital allocation processes. These are conversations that we have with the leadership team around how we must continue to make sure we’re making the right decisions – and even using technology to help support those decisions, testing some of our investments using AI to see how they compare to things we have done in the past.
Operational excellence is another area we mentioned. Our operating efficiency in our upstream facilities is in the range of about 95%, and we’re continuing to look at how we improve it. Every single percentage point in operating efficiency, especially given today’s macro environment, is extremely valuable to us as a company, and technology is a big part of how we push on that. Availability in our refineries is also key for us. So on both our operating efficiency and our refining availability, we’re looking at the 94–95% range and continuing to improve through great people, great processes and the best application of technology.
Then there are our downstream businesses and our retail businesses, where we serve our customers, either on the retail side or the aviation side – again, a really important part of our business.
Finally, trading is a massive part of bp’s activity set and operations, and another big place where people, processes, technology and the right assets have a big impact.
How big an opportunity is AI for bp and how is the company approaching it?
Emeka Emembolu: It’s a massive opportunity for us at bp. You hear about AI externally all the time – it’s a constant topic of conversation. But we take AI very seriously as an opportunity to continue to build the competitive advantage that I talked about. It’s probably worth me talking about how we approach it, and then we can talk about what we actually do with it as part of that approach.
We have a pretty clear framework that we use to approach AI. The first part of that framework is business focus. Everything that we do in technology, especially with AI, needs to have a business purpose. We don’t want to bring in technology or partners, or try to do things, that are not addressing a business need or a business problem. We can talk about what some of those business problems look like and flesh that out a bit more.
The second is the foundations, and I’d put those in three broad buckets. One is data. Data is a massive foundation that we need to have in place for AI – people sometimes say data is the fuel for AI. Another is the partnerships, and we have some really key partnerships. AWS and Microsoft are key to us as cloud providers, and they help us with other capabilities. Databricks helps with our Unified Data Platform that we’re building out across the group, with Palantir helping with the data insights for that platform. I mentioned Nvidia, who we work with on the chip front to build out our HPC, which we use for both seismic and other forms of machine learning, AI and compute. We have Infosys and Accenture, whom we work really closely with as partners. SAP is another important one for us. And finally, Anthropic, on the AI side, as one of the foundational model providers. That’s our core ecosystem that sits inside this foundational pillar.
The third pillar is trust. This is around how we make sure we can trust what we’re getting out of AI, and how we monitor the cost, because we want to make sure everybody knows how much they’re spending. You hear all the horror stories about people overspending on AI. We want to make sure that people can use what we have within bp with confidence. So trust is really important – safety, regulatory requirements and even cyber safety, and how we use AI in that space to make sure that we are running the company in the right way.
And finally, the last area is fluency. How do we make sure we’re training everybody in the company, from the executive team all the way through the organisation, in terms of building fluency with AI? These are some of the key pillars of how we approach AI at bp.
What are bp’s main AI and technology use cases across its upstream, downstream and trading businesses?
Emeka Emembolu: I’ll start in the upstream. When you go to the upstream, you always have to start with the rock – subsurface is where it all begins. A key starting point is geophysics and high-performance computing. What we’ve done there is significantly increase the capacity of our high-performance computing system based in Houston. That has unlocked the ability to process seismic faster than we ever have before, but also to solve problems we were never able to solve when we had the reduced capacity. The outcome is that our explorationists and subsurface teams have these tools that they can use to drive the exploration success that we’ve seen. This has supported some of what we have done in Brazil, in Namibia and in the Gulf of America. These are some of the things that we’ve been able to do with technology supporting our geoscientists.
Moving on from there, reservoir simulation is another area where we’re investing. We have selected a new simulation system as a company, and AI is front and centre in that new system, as is the compute in our high-performance computing centre.
Moving into production, production optimisation has been bread and butter for us for a number of years. The Palantir Foundry system has been critical to some of the percentage improvements that we’ve made in operating efficiency over a number of years. It’s helped us generate significant additional value for our company in terms of free cash, and obviously for shareholders directly.
If I move to the downstream, there are probably two areas worth touching on. First, refining. Over the years we have not invested as much in our refining digital systems as we should have done, but last year we made a course correction and we’re investing more. We’re moving from much older ways of operating to a more digitally enabled way of running that business. Brenda Stout, our senior vice president for refining, is a massive advocate for what we can do with digital systems and has been a huge driving force in making sure we’re getting the right outcomes. We’ve just implemented a new system in our Castellón refinery, and we’re going to be expanding it to the rest of the refineries. It helps the operators on the front line manage their activities day to day and be safer, and it connects and integrates a whole number of previously non-integrated systems in a way we haven’t had before.
That’s refining. In our retail business, something our retail team are particularly proud of is our earnifyTM platform, which we operate in the United States. This is our loyalty system. Listeners in the United States would actually see that our bp earnifyTM platform is one of the top 10 apps in navigation. It’s a platform we use to make sure people are informed about our loyalty offer, but behind the scenes there’s a lot of technology – we used AI to build it, and AI is very evidently powering it to help make sure our customers get the best deal. That’s another area we’re really stepping into, and we’re looking to unify and harmonise our technology offer for our retail business. So this has been a big deal.
If I move to trading, that’s another really important part of our company. Trading is one of the points at which we integrate both the upstream and the downstream to maximise value and give additional returns for our company. We’re making a significant investment in our trading technology stack in order to get it to a different place.
As an example of what we’re doing in trading, just prior to coming into this meeting, I met with one of our traders, who shared with us how he’s been integrating a number of different data sources. We’re building AI advisers for a number of our traders. We’ve rolled them out in parts of trading, and they’re going to be expanding to the rest of trading over the next two years. They provide the type of access traders have never had before to all of our data that sits in trading, and link data from beyond the company into that trading system – using AI to help traders make better trades and faster trades, and to be more competitive in the market.
So we’re really excited about a lot of what we’re doing in the technology space, all underpinned by a data platform that we’re building with Databricks, which we call our Unified Data Platform.
How is bp combining robotics and AI across its operations?
Emeka Emembolu: Thanks for asking, because this is an area that I feel really excited and really confident about. Robotics is one of the key strategic themes in our upstream technology roadmap. Across the company, we have used robots in one form or another – either drones or crawlers – and we use them regularly in a number of our assets. But we haven’t really put together an integrated platform that looks end to end at how these robots interact with the physical world and how that data flows into our systems in a fully effective way.
To build this ecosystem out, we are concentrating on one of our terminals, Sangachal in Azerbaijan, and building out our full robotic system there. We’re using it as a flagship location to build out our fully integrated system. The intent is then to expand from there and roll out this new platform across the rest of the group over the next few years.
The team out at Sangachal have been amazing in terms of pulling on the technology, adopting it, testing it, and seeing what the limits are and what’s possible. A great example is something we showcased in Azerbaijan back in July: one of the drones that we demoed could actually be used to repair spot defects in some of our pipelines or tanks. What this unlocks is the ability for people, or even machines, to go around the site and identify a defect, and then the local team can easily use these drones to do repairs. So it’s moving from just observation to actually undertaking tasks. In time, we hope to get to the point where those tasks can be done autonomously, so that we can dispatch drones to do fixes and repairs on our sites.
So I think the future for robotics in the oil and gas industry is extremely bright and really exciting.
This transcript has been lightly edited for clarity and readability. Verbal fillers, false starts and minor repetitions have been removed from the interviewee’s responses only. Punctuation, spelling and formatting have also been standardised in line with Edison house style. No substantive changes have been made to the meaning of the discussion.
