Applied Computing, a London-based startup, is betting that the future of heavy industry lies in a single, plant-scale AI brain. The company is building what it calls a foundation model for oil, gas, refining and petrochemical facilities, designed to understand every pipe, pump and valve in real time.

The startup has raised a 20 million dollar Series A round led by engineering giant KBR, with Databricks Ventures also investing. Founded in 2023, Applied Computing is targeting a long-standing problem in complex industrial sites: operators are drowning in data but starved of insight. A large refinery or petrochemical plant can host thousands of sensors tracking temperature, pressure, flow, vibration and more, yet only a small fraction of that information is used in day-to-day decisions.

Chief executive and co-founder Callum Adamson says the challenge is not collecting data but making disparate streams work together fast enough to matter. Operators must reconcile sensor readings with engineering documentation and the underlying physics and chemistry of their processes. Doing that manually or with traditional software can take days or weeks, slowing investigations into anomalies and limiting the ability to test new operating strategies.

Applied Computing’s answer is Orbital, a foundation model that blends three approaches: time series analysis of sensor data, physics-based modeling of industrial processes and a language model that can interpret documentation and operator input. Instead of predicting the next word, Orbital predicts the evolving state of an entire facility, taking into account equipment constraints and human actions.

The company says Orbital can flag anomalies, trace likely root causes and simulate potential fixes across the plant within minutes. Investigations that once stretched over days can be compressed into seconds, Adamson claims, allowing operators to cut energy use, avoid unplanned downtime and maintain output.

Applied Computing reports that it has moved from stealth mode to double-digit millions in annual recurring revenue in under 18 months. Orbital is already deployed with large, publicly listed companies in upstream oil and gas, downstream refining and petrochemicals, though the startup has not disclosed customer names.

KBR has integrated Orbital into its INSITE 3.0 digital platform and is using it in ammonia production projects. Applied Computing is also working with Indian energy firm Wipro, a major United States upstream operator and is preparing a partnership with a European oil major. The new funding will support international expansion, research and engineering hires and further deployments across global energy hubs, including a growing presence in Houston and plans for the Middle East.