The energy giant’s EVP of innovation and engineering, David Eyton, describes its formula for anticipating — and exploiting — the most disruptive digital tech.When it comes to technology strategy, multinational energy company BP plays by the classic ice hockey maxim, ‘Skate to where the puck is going to be, not where it is.’To support that goal of positioning itself today to ensure competitive edge in the future, the UK-headquartered company operates an elite technology unit, the Digital Innovation Organization (DIO), that sits apart from the teams that run IT within business units such as Upstream, Downstream, Alternative Energy and Lubricants.DIO is all about investigating far-horizon technologies that have the potential to be disruptive, working closely with other BP teams, academic institutions and start-ups. As David Eyton BP’s EVP of innovation and engineering, outlined in an interview at The Economist Innovation Summit in London: “We track 425 different technologies that are evolving. We boil that down to what is possible in the future and then feed those into the strategy-making process of the company.” And that means knowing when to invest in a technology, even as it is taking shape, but previous and on-going focuses have been on digital twins, distributed ledgers, advanced AI and quantum computing.“If you just base technology strategy on [what is feasible] today then you’d say it’s too expensive to invest,” said Eyton. Electric car engines provide him with a good example. They might currently be priced at a premium, but the downward slope of battery costs means they are going to be much more cost-effective than internal combustion engines in the longer term,” he added.
David Eyton, EVP of innovation and engineering, BP
Having identified valuable innovations, DIO runs proof-of-concept and pilot projects with relevant business teams, and then helps scale up the ones with the greatest potential. But the focus is distinctly on technology for BP’s frontline.
“Most of the company is out there, working in refineries and petrochemical plants, in biofuels, wind and solar farms, and in gasoline [retail] or electric charging systems. That’s where there’s an interface with the customers, and that’s where we seek to innovate — improving the cost of doing business and the quality of service,” he explained.
Subsurface intelligenceEyton points to one project in that vein known by the nickname ‘Sandy.’
This is a powerful ‘physics-based AI platform’ and digital assistant designed to unlock critical data for BP’s subsurface engineers at a much-accelerated pace. BP experts feed the platform with vast quantities of geology, geophysics, reservoir and historic project information — often in radically different formats.
Sandy (orginally developed by Houston-based Belmont Technology, in which BP is an investor) intuitively groups that information together, identifying new connections and workflows, and creating a knowledge-graph of BP’s subsurface assets. Subsurface engineers can then interrogate the data, asking the knowledge-graph specific questions in natural language, such as ‘What factors control production in the Chirag field [in the Caspian Sea]?’ or ‘What is the average porosity of the Miocene reservoirs?’ The technology then uses AI neural networks to interpret results and perform rapid simulations, helping engineers to quickly understand situations, generate novel ideas and make better-informed decisions.
‘Talk to Sandy’ — Physics-based AI and smart assistant developed by BP-based Belmont Technology
The upshot is an acceleration of project lifecycles, from exploration through to reservoir modeling. The target for the system is a 90%-time reduction in data collection, interpretation and simulation. And BP already says Sandy has helped solve some simulations 10,000 times faster.
What is important about the system, said Eyton, is that ability to aggregate and interrogate hugely complex information. “Historically, you’ve had people’s actions constrained by what they can get their minds around. We have geophysicists working on interpretation of seismics, who hand off data to geologists working on the way rocks are laid down, who hand off to petro-physicists [looking at] the character of the rock, who in turn hand off to drilling engineers,” he said.
The need to apply different expertise at different stages of projects means there is no universal appreciation of the range of possible outcomes. “But if you can interrogate everything in one go using AI that’s been informed by all your experts, you can immediately come up with a most likely scenario, as well as all of the other scenarios that are possible” he said.
As that just amplifies Eyton’s core argument for “a strategy that says, ‘we’ll invest in something because of where it’s going rather than where it is today.’”