London-based startup Applied Computing has raised $20 million to build an AI system called Orbital, designed specifically for the oil, gas, and petrochemical industry. The funding round was led by KBR, an engineering company, with Databricks Ventures also participating. TechCrunch

Founded in 2023 by CEO Callum Adamson, the company has been publicly operating for less than 18 months and says it is already bringing in double-digit millions in yearly revenue. Orbital is currently being used by large, publicly traded oil, gas, and petrochemical companies, though Applied Computing has not named them.

Orbital works by combining three different kinds of AI into one system. The first looks at patterns in data collected over time, like temperature or pressure readings from sensors. The second understands the laws of physics and chemistry that govern how a plant actually works. The third is a language model, which lets people ask questions and get answers in plain English. Together, these let plant workers simulate how a change in one part of a facility would affect everything else. Applied Computing claims Orbital can spot problems, figure out what caused them, and test whether a proposed fix might create new problems elsewhere, all within minutes instead of the hours or days that older methods typically take.

Adamson says the problem comes down to wasted data. He states that facilities currently make operating decisions using less than 8% of the data available to them. The rest, collected by sensors throughout a plant, goes unused. Orbital is designed to take in all that data and find useful patterns in it.

KBR's investment includes a seat on Applied Computing's board. Investing.com KBR has already built Orbital into its own digital platform for energy projects and is using it for ammonia production. Wipro is also listed as a partner, though the details of that relationship have not been shared.

On the customer side, Applied Computing says it is working with a major U.S. oil and gas producer and plans to announce a partnership with a European oil major in the coming weeks.

The design choice of using three models together rather than a single AI is what makes Orbital technically interesting. A language model on its own is good at understanding text but cannot reliably account for the physical and chemical rules that govern an industrial plant. By pairing a model that reads sensor data with one that enforces physics and chemistry rules, and then adding a language model so people can interact with it conversationally, Orbital tries to combine the strengths of engineering simulation with the pattern-finding abilities that modern AI is good at.

The broader context is that AI has already proven useful for generating text, code, and images. Using it in industrial plants is a much harder problem, because a wrong recommendation could lead to equipment damage or a safety incident rather than just a bad piece of text. The way Applied Computing has designed Orbital, tying the language model to physical rules instead of letting it make free-form suggestions, is a sensible response to that risk.

A few things are worth noting. The claim that Orbital can do its work in minutes has not been independently verified at large scale. The company's revenue figures and customer list, while reportedly real, are self-reported. That said, the fact that both KBR and Databricks Ventures have put money in and integrated the product suggests they see real technical and commercial promise.

For plant operators, the potential benefits are clear: catching problems faster, understanding their causes more quickly, and testing fixes in a simulation before trying them on a live facility. Whether Orbital works well across many different types of plants and processes will determine whether the company's fast revenue growth reflects lasting demand or just early excitement. The upcoming European oil major announcement will be worth watching.