BP: an energy operations digital twin built on Palantir Ontology

BP has used Palantir software since 2014 to build a digital twin of its oil and gas production that combines real-time data from more than 2 million sensors. In September 2024 the two companies signed a new five-year agreement that adds Palantir AIP’s large language model capabilities.

Summary

Item	Detail
Company	BP
Industry	Energy
Use cases	Predictive maintenance, data-driven decisions, business intelligence
Functions	Production, strategy
Reported result	No financial or production figures disclosed; public facts are 2 million+ connected sensors, a 10-year partnership and a 5-year renewal
Moments rating	S-tier quick win
AI use-case score	90

Why it matters

BP spent years joining sensor data and asset models into one operational view before layering LLM-based recommendations on top. That order, data foundation first and AI suggestions second, is the part worth copying.

Case details

Background

According to Offshore Technology (September 10, 2024, republished by Yahoo Finance), Palantir software has supported BP’s operations since 2014. Deployments include offshore oil platforms in the North Sea and the Gulf of Mexico and the Khazzan gas field in Oman, supporting BP’s digital transformation and the optimization of its oil and gas production. After roughly a decade of working together, the two companies signed a new five-year strategic agreement.

What they did

The core of the work is a model-based digital twin of BP’s oil and gas production: dynamic digital models of physical assets combined with real-time data from more than 2 million sensors, with Palantir software providing a single operational view. The new agreement brings in Palantir AIP so BP can use large language models to speed up and improve human decision-making. AIP suggests actions based on automated data analysis and includes safeguards against AI errors, the so-called “hallucinations”.

Results and data

The report does not disclose cash flow, output or cost savings. The only scale figure given is the 2 million-plus sensors. Sunjay Pandey, BP’s senior vice president of digital delivery, said advanced digital twin simulation helps BP monitor and optimize production safely. Matthew Babin, Palantir’s head of energy and natural resources, said AIP helps accelerate human decisions.

What to take from it

Build the unified view of sensors and asset models first, then add LLM recommendations. Without a reliable data foundation, model suggestions are hard to act on. Having the AI recommend while people decide is also a sensible split of work in high-risk production settings.

Limits

The public reporting contains no outcome data, so return on investment cannot be estimated from it. This path rests on a ten-year partnership and a very large sensor deployment, which means long timelines and heavy investment. Energy companies in China would also need to account for rules on critical information infrastructure and cross-border data transfer.

Source

Original source: Yahoo Finance · BP extends partnership with Palantir

This is an industry research entry: a publicly documented third-party AI case, not a Moments client case. For projects Moments delivered itself, see client cases.