Transcending Plant Potential: Integrating Deep Learning with the Foundation Process Model™ in the Optimizing Brain™ Solution
By Gil Cohen, CEO and Co-Founder of Imubit
Imubit’s Optimizing Brain™ Solution leads the market in Closed Loop AI Optimization, transforming engineers into AI pioneers with its Foundation Process Model™ and reinforcement learning. Similar to how ChatGPT, the pioneering Foundation Model, was trained on vast datasets to power diverse applications, our model integrates extensive process data with engineering expertise. This evergreen asset continuously updates with live process data, offering clients a versatile tool for optimizing operations, training teams, and diagnosing process issues.
The world of artificial intelligence has been transformed by Foundation Models—large, adaptable models like OpenAI’s ChatGPT. Trained on vast datasets through the combined efforts of data scientists, engineers, and immense computing power, these models have revolutionized AI’s ability to tackle complex problems—similar advancements are now being made in the industrial world.
At Imubit, we are witnessing a step change within industrial plants, and we’ve coined this innovation the Foundation Process Model. This new concept, built from deep learning applied to large sets of process data, is poised to redefine how plants operate and optimize performance.
From Local Solutions to a Universal Process Model
Historically, industrial plants relied on local models and logic to address specific, isolated problems. These models ranged from Linear Programming (LP) to first-principle process models, Advanced Process Control (APC) applications, scheduling models, and operator training systems. While each served its purpose, they often required extensive manual input, fine-tuning, and optimization specific to the task at hand.
Now, we are seeing a shift. Much like how foundation models in AI were trained on broad datasets for wide applicability, plants are training deep learning models on large amounts of process data, investing multidisciplinary efforts to create a single model that is applied to a broad set of applications.
Building the Foundation Process Model: A Collaborative Effort
Developing a Foundation Process Model™ requires a collaborative effort between the various technical teams at the plant. Engineers start by investing in building models based on years of process data, carefully selecting and cleaning data, and integrating their deep knowledge of the process and the appropriate first principles.
Once trained, these models are validated by engineers, who analyze model accuracy, error margins, and how well the model captures first-principle relationships. The result is a model that becomes a valuable asset, evergreen through training on fresh plant data, and reused to solve various problems within the plant’s operations.
Real-World Use Cases
At our recent Imubit Transcend AI Summit and Training, our customers shared how they are already using Imubit Foundation Process Models in innovative ways. For example, Big West Oil’s Operations Supervisor Bobby Hammer presented how their operators and engineers are learning optimization strategies not through generic training simulators, but with specific models based on Big West’s data and unique challenges.
Kevin Daines, Engineering Technical Specialist from HF Sinclair shared how LP modelers are now able to update shift vectors with greater frequency—potentially from annual to weekly—allowing for more responsive and efficient operations. In Europe, one of our clients is using the Imbuit Foundation Process Model to track catalyst degradation over years in a conversion unit, ensuring more precise and timely directives to operations teams.
The Future of Process Optimization
The potential applications of the Foundation Process Model are far-reaching. Beyond day-to-day operations, companies are using these models to justify capital projects through before-and-after gain analysis, to fine-tune PID loops with unprecedented accuracy, and to explore countless other opportunities.
By eliminating the need for approximation and relying on broad, deep learning models, the Foundation Process Model represents a fundamental shift in how plants operate. As companies continue to adopt and refine these models, we expect them to become indispensable assets for plants aiming to stay ahead in an increasingly complex and competitive landscape.
At Imubit, we are excited to be at the forefront of this revolution, helping industrial plants unlock the full potential of their process data and achieve new levels of efficiency and optimization.
Unlock the power of the Foundation Process Model™ with the Imubit Optimizing Brain™ Solution. Leverage data-driven models across your entire operation, break down plant silos, and empower every technical team member to excel in an AI-driven future.
