Applied Computing debuts physics-based AI for energy plant operations
Applied Computing has released Orbital 2.0, an AI platform that utilizes a 'world model' to simulate plant operations and physical cause-and-effect relationships.
At the ADIPEC conference in Abu Dhabi, Applied Computing unveiled Orbital 2.0, an update to its AI software for energy and petrochemical facilities. Unlike its predecessor, which relied on pattern recognition to perform routine tasks, the new architecture models the physical interactions within a plant, such as heat and mass transfer, to provide operational insights.
The system is designed to allow engineers to visualize the plant-wide impact of potential decisions. Each client operates a private version of the model, ensuring that proprietary data remains within their own infrastructure. The new version introduces self-service capabilities, allowing users to generate automated reporting tools and dashboards via prompts.
This shift in architecture reduces the deployment timeline for new customers from several weeks to a few days by minimizing the need for custom tuning by the vendor. The company, which secured $20 million in Series A funding in July 2026 led by KBR, aims to provide operators with a tool that continuously learns from both historical data and direct operator feedback.