Applied Computing's AI Model Revolutionizes Oil and Gas Operations

Applied Computing is building a comprehensive AI foundation model for the oil, gas, and petrochemical industries, aiming to unify fragmented data and significantly enhance operational efficiency.

Unifying Fragmented Industrial Data with AI

London-based startup Applied Computing recently completed a significant Series A funding round, securing $20 million. The investment was led by engineering giant KBR, with participation from Databricks Ventures. This capital infusion will support the company's mission to develop a pioneering foundation AI model specifically tailored for the intricate operations of the oil, gas, and petrochemical sectors.

The challenge within these industries stems from the sheer volume and diversity of data generated across a single facility. Thousands of sensors continuously monitor a vast array of parameters, from temperature and pressure to fluid velocity and viscosity. Despite this rich data environment, many facilities currently utilize less than ten percent of their available data for operational decision-making. The core issue, according to Callum Adamson, co-founder and CEO of Applied Computing, is not a lack of data collection but rather the difficulty in integrating disparate data streams. Sensor readings, engineering documentation, and complex physical and chemical models often operate in silos, hindering real-time analysis and predictive capabilities.

Adamson emphasizes that the critical innovation lies in enabling these distinct data sources to communicate seamlessly and instantaneously. This integration is paramount for transforming raw data into actionable insights, allowing operators to make informed decisions that can optimize performance and prevent costly failures.

Orbital: A Multi-Modal Foundation Model

Applied Computing's flagship product, Orbital, differentiates itself from conventional large language models. While LLMs primarily focus on predicting text sequences, Orbital is engineered to predict the dynamic state of an entire industrial facility. It achieves this by integrating three key components: a time-series model for analyzing sequential sensor data, a physics-based model incorporating foundational chemical and physical principles, and a language model to understand context from operational logs and documentation. This multi-modal approach allows Orbital to process sensor data while simultaneously accounting for the inherent physics and chemistry governing the processes, as well as recognizing equipment constraints and human operator activities within the plant.

A significant feature of Orbital is its simulation capability. It empowers technicians to run 'what-if' scenarios, modeling the potential impact of a change in one part of a facility on the entire system. This predictive modeling is crucial for understanding systemic effects before implementing actual modifications, thereby mitigating risks and optimizing outcomes.

Accelerating Decision-Making and Operational Efficiency

The most compelling advantage offered by Applied Computing’s Orbital platform is speed. The company claims Orbital can identify anomalies, diagnose their root causes, and simulate the effects of proposed solutions across the entire facility in a matter of minutes. This drastically reduces the time required for investigations that previously could take days or even weeks. By accelerating this process, operations can maintain consistent output and minimize energy consumption effectively.

This promise of enhanced speed and efficiency has garnered considerable traction. Applied Computing has moved from a stealth phase to generating double-digit millions in annual recurring revenue in less than 18 months. While specific customer numbers were not disclosed, Adamson confirmed that Orbital is already deployed at multiple large, publicly listed upstream oil and gas, downstream refining, and petrochemical companies. Partnerships include an integration with KBR’s INSITE 3.0 digital platform for energy projects and collaboration with Indian energy company Wipro. The company is also working with a major US upstream operator and anticipates announcing a partnership with a prominent European oil major in the near future.

Navigating a Competitive Landscape

Applied Computing operates in a market that already contains established industrial software providers and emerging AI specialists. Companies such as AspenTech offer simulation and AI-driven modeling solutions for various energy operations. AVEVA provides physics-based process simulation, optimization, and 'what-if' modeling tools for industrial plants. Furthermore, companies like Cognite and Seeq focus on the data layer, helping facilities analyze industrial data and apply AI to workflow design.

Adamson asserts that Applied Computing’s competitive edge lies not in proprietary access to industrial data or process knowledge, but in its ability to attract and retain top-tier AI researchers. He argues that constructing a model capable of Orbital’s comprehensive functionality is primarily an AI research challenge, not merely a data or energy industry problem. He suggests that leading AI researchers are more likely to be drawn to cutting-edge AI development environments than traditional energy companies.

The real-world operational data obtained through Orbital’s deployments further reinforces its competitive moat. Adamson notes that data from refineries and other energy facilities is rarely public, and simulated data often fails to capture the full complexity of live plant conditions. The partnership with KBR is also strategic, providing not only capital but also access to valuable operational data and industry expertise, which facilitates introductions to new potential clients.

Future Expansion Plans

The recently secured $20 million in funding is earmarked for several key expansion initiatives. Applied Computing plans to broaden its international presence, significantly invest in research and engineering talent, and explore further deployments with energy clients globally. The company has already established an office in Houston, complementing its London headquarters and Bengaluru operational hub. This move brings Applied Computing closer to its existing North American customer base. Expansion into the Middle East is also reportedly underway, as detailed in a recent TechCrunch report.

Why it matters

This development is significant for the infrastructure and industrial sectors. The oil, gas, and petrochemical industries are foundational to global energy and material supply chains. By providing a unified AI model that can process diverse data streams, Applied Computing aims to usher in a new era of predictive maintenance, operational efficiency, and risk mitigation. This directly impacts data center operations by requiring robust, low-latency infrastructure to handle the real-time processing of vast industrial sensor data. For telecommunication providers, the increasing demand for seamless data flow between geographically dispersed industrial assets and central AI processing hubs highlights the need for advanced network capabilities. Technicians in these facilities will find their roles augmented by AI, shifting from reactive problem-solving to proactive, data-driven decision-making, ensuring safer, more efficient, and more sustainable operations.