Energy plants collect enormous volumes of operational data. The harder problem is turning it into decisions operators can trust.
London-based Applied Computing has raised $20 million in Series A funding to expand Orbital, its foundation AI platform for oil, gas, refining, and petrochemical operations.
The round was led by engineering and technology company KBR, with participation from Databricks Ventures. Applied Computing will use the funding to support international expansion, product research, and wider deployment across industrial sites. It has also announced a new office in Houston, placing the company closer to major US energy operators.
Connecting plant data, physics, and language
Founded in 2023 by Callum Adamson and Samyakh Tukra, Applied Computing is building AI specifically for complex energy facilities.
Its Orbital platform combines time-series sensor data, engineering physics, and language models. The system is designed to help operators understand plant behavior, test operational decisions, and identify opportunities to improve efficiency, reliability, and emissions performance.
Applied Computing says industrial facilities may act on less than 8% of the data they collect. A refinery can generate thousands of readings across temperature, pressure, flow, velocity, and viscosity, but those data streams are often separated across legacy systems and specialist teams.
Orbital’s role is to connect those inputs rather than replace plant engineers. Its physics layer checks whether an AI-generated recommendation is technically feasible, while its language interface makes the analysis easier for engineers to explore.
Why energy needs specialized foundation models
General-purpose AI models can summarize documents or answer technical questions, but operating a refinery requires an understanding of physical constraints, equipment behavior, and changing process conditions.
A recommendation that looks efficient in software may be unsafe or impossible inside a live plant.
Applied Computing is therefore pursuing a vertical foundation-model strategy: train and operate AI around one industry’s data, language, and physical rules. The company says its technology is already being used in large refinery environments and is designed to support real-time operational decisions.
The market signal
Industrial AI is moving beyond dashboards and predictive-maintenance alerts.
The next layer is plant-wide intelligence that can combine sensor readings, engineering models, maintenance information, and operator knowledge within one system.
Applied Computing’s challenge will be proving that Orbital can deliver measurable improvements without adding operational risk. Energy operators will expect clear audit trails, strong cybersecurity, and recommendations that engineers can independently verify.
The $20 million round also highlights growing investor interest in specialized AI systems built around physical industries. The market is testing whether domain-specific models can create stronger commercial value than general AI tools adapted after deployment.
Applied Computing is betting that energy companies do not need another isolated analytics product. They need an AI layer capable of understanding how the entire plant works.
Source : Tech.eu
