Imubit

Imubit develops Closed-Loop Physical AI for process industries, including refining, chemicals and polymers, cement, and mining and minerals. Founded in 2016 and headquartered in Houston, Texas, the company says its platform is designed to help industrial operators address gaps between operational plans and how facilities actually run. It combines analysis of plant-wide objectives and operating decisions with process models and controls that act within limits established by plant engineers.

The platform is organized around Map, Model, and Control. Map brings together planning information and plant operations to identify and prioritize differences between intended and actual performance, tracing them to recurring decisions, disturbances, and outcomes. Model uses plant-specific operating history, deep-learning methods, and first-principles guidance to represent cause and effect in process units. Imubit says process engineers can build and work with these models through its DLPC Builder without needing a machine-learning background.

Control applies the models to selected recurring operating decisions. The company describes its controllers as using reinforcement learning during development and producing deterministic actions in operation: the same inputs yield the same move. Controllers can send setpoints through an existing advanced process control system or directly to a distributed control system, operating within limits set by plant engineers. Imubit states that operators retain authority and can take over at any time.

Imubit serves process manufacturers and operators whose facilities include refineries, chemical and polymer plants, cement kilns, and mineral-processing operations. Its stated use cases include improving process performance and energy efficiency while working with existing plant control infrastructure. The company reports more than 100 closed-loop controllers deployed worldwide, over 300 applications deployed across its portfolio, and more than seven years of closed-loop engagement in production. These are company-reported figures.

The company’s origins and leadership are tied to engineering and industrial technology. Its official company information identifies Gil Cohen as CEO and co-founder, and says he has led Imubit since its founding. It also identifies Prof. Nadav Cohen as co-founder, president, and chief technology officer. Imubit reports thirteen approved U.S. patents. Its Houston-based business operates internationally, with deployments described by the company as worldwide.

Imubit’s careers information describes a team spanning machine-learning research, process and control engineering, industry expertise, and enterprise software engineering. The company presents its work as closely connected to the physical operations of customer plants, where software must operate within safety and process constraints. Its published working principles include teamwork, passion for collaboration and learning, humility and iteration, individual ownership, purpose-driven leadership, and commitment to its mission. It also describes delivery teams working with customer plant personnel from initial data packages through deployment, with the aim of enabling customers’ engineers to operate and maintain their models.