Applied Computing raises $20M Series A to scale Orbital AI for oil, gas and petrochemicals
Applied Computing raises $20M to scale Orbital, an AI model for oil, gas and petrochemical plants that fuses time-series data, physics and language models.
Applied Computing announced a $20 million Series A investment on Thursday to accelerate deployment of Orbital, its foundation AI model for the oil, gas and petrochemical industry. The London-based startup says Orbital fuses time-series sensor data with physics-aware models and language understanding to predict facility state and speed incident investigations. The round was led by engineering firm KBR with participation from Databricks Ventures, and the company plans to use the proceeds to expand internationally and deepen research and engineering capacity.
Funding and strategic backers
KBR led the $20 million Series A, signaling a strategic alignment between the engineering contractor and Applied Computing’s product roadmap. Databricks Ventures participated in the round, and the startup listed partnerships with Wipro and integrations into KBR’s INSITE 3.0 platform as part of its commercial momentum.
Applied Computing said the financing will fund international expansion, hiring for research and engineering roles, and further deployments with energy clients. The company also opened an office in Houston to support North American customers, complementing its London headquarters and operational hub in Bengaluru.
How Orbital combines multiple AI approaches
Orbital is described by the company as a hybrid foundation model that blends time-series forecasting, physics-based simulation and language modeling. The approach aims to map thousands of disparate sensor streams to engineering documentation and the underlying chemistry and physics that govern plant behavior.
Rather than predicting the next token like large language models, Orbital reportedly models the state of equipment and processes, enabling scenario simulations and root-cause analyses that take physical constraints into account. Applied Computing positions this architecture as tailored to the complex operational dynamics of refineries, upstream platforms and petrochemical plants.
Claims on speed and operational impact
Applied Computing says Orbital can detect anomalies, trace their probable causes and simulate proposed fixes within minutes, compressing investigative timelines that previously took days or weeks. The company estimates that operators typically use less than 8% of available sensor data today, and argues faster synthesis of data and physics can reduce energy use while preserving output.
The startup claims double-digit millions in annual recurring revenue within 18 months of commercial activity, and says Orbital is already in use at several large, publicly listed energy companies. Applied Computing declined to disclose exact customer counts but emphasized deployments across upstream oil and gas, downstream refining and petrochemicals.
Commercial partnerships and deployment use cases
Applied Computing highlighted KBR’s use of Orbital within INSITE 3.0 and cited a specific deployment supporting ammonia production. The company also confirmed work with a “major U.S. upstream operator” and said a partnership with a European oil major is expected to be announced in the coming weeks.
The Houston office is intended to bring the startup closer to North American customers and pipeline opportunities, while an expansion into the Middle East is under consideration. Partnerships with systems integrators and engineering firms are central to Applied Computing’s go-to-market strategy, according to the company.
Competitive landscape and technical differentiation
The market for industrial AI and process simulation is crowded, with established vendors such as AspenTech and AVEVA offering physics-based process models and simulation suites. Data-focused platforms like Cognite and Seeq also target the same operational analytics layer by helping companies organize and analyze industrial data.
Applied Computing argues its competitive moat stems from assembling AI researchers and engineering talent to build a model that natively combines time-series, physics and language capabilities. The company further points to access to operational data through partner deployments as a source of training signal that simulated datasets cannot fully replicate.
Risks, adoption hurdles and enterprise dynamics
Despite the funding and early traction, Applied Computing must contend with entrenched industrial software suppliers, lengthy procurement cycles and the challenge of integrating with heterogeneous sensor and control systems. Industrial customers often require extended validation, traceability and alignment with safety and regulatory practices before adopting new models in mission-critical environments.
Applied Computing’s commercialization relies on its ability to demonstrate reliable predictions and clear return on investment in live operations, and on partners like KBR to provide domain expertise and customer introductions. The company will need to scale operations while maintaining model performance across diverse facility types and geographies.
Applied Computing plans to deploy the new capital to accelerate product development, broaden its customer footprint and recruit research engineers, with particular emphasis on North America and the Middle East as growth targets. The company says Orbital’s blend of data, physics and language modeling is designed to address what it calls an “AI problem” at the intersection of industrial operations and advanced modeling.
