Innovation to Support Your Business and Creativity

You are creative and innovative, and Intellect supports your abilities to compose outstanding, industry-leading solutions. There are no imposed limits.  You can craft solutions that will offer substantial benefits to your company’s bottom line, leading the industry in innovation.  All Intellect solutions are custom, though there are a series of successes experienced by those who have come before you.  In oil and gas, virtual metering, production optimization has been deployed worldwide on large scales.  In mining and metals, in aerospace, pulp and paper, consumer products, discrete assembly or any combination there of.  Most industrial customers have hybrid operations and you likely do to.

Innovative Machine Learning

Self-Optimizing Modeling

In 1995, we invented Genetically Optimized Neural Networks to create predictive “Machine Learning” models (ML).  The Genetic Algorithms search through combinations of inputs, model types, and model internal structures to find models that work well across one modeling and two validation datasets.  The best models are kept, and a committee of high performers is formed.  The estimates from this committee are merged to create top-performing predictions.  The models built include predictive regression, probabilistic, generalized regression, and other model types. These systems of models can be stored on disk to use in on-line predictors, getting data from DCSs, SCADA systems, data historians, SQL databases, etc. and write the predictions to our onboard historian, or back to your systems.

Human Attended or Autonomous, Often Both

These models can be built by a modeling expert in our “Expert” tool or autonomously by our Intellect Server.  Rebuilding can be done “lights out” based on almost any criteria, including model performance or loss of input sensors.  Predictors run automatically and different data, handled differently are often used.  Auto-adaptation of performance is the norm and limits can be imposed when data goes “awry”.

Intellect is Highly Scalable

Intellect scales to meet your needs no matter where you are on your journey. From small-scale proof of concept evaluations to optimizing operations nationwide across multiple data centers, Intellect is ready when you are to grow.

The journey begins with a straightforward installation onto individual computers for small-scale operations. As you progress, Intellect is integrated into clusters using Consul, a microservice discovery tool. We containerize Intellect and run multiple instances to facilitate deployment and manage higher-capacity computing. The next step involves orchestrating these containers using Docker Swarm or Kubernetes for enhanced capacity and fault-tolerance management. And yes, we even span across multiple data centers for optimizing all assets simultaneously across an entire nation.

Intellect Technical Ladder

What sets us apart is our ability to operate on-premises, coexisting with your distributed control systems, programmable logic controllers, and data historians. We bridge the gap between your control and office networks, spanning the “DMZ”.

We’ve successfully executed these systems multiple times worldwide, demonstrating Intellect’s capacity and reliability.

Task-Oriented, Message Passing Architecture

Intellect Is Highly Extensible

Intellect’s operations are made of independent tasks that are “wired together” with interprocess messaging, enabling solutions of any type and capabilities of any capacity. Multiple tasks work together as a team in almost every process in an Intellect solution to detect operating states, make predictions, conduct optimizations, and more.  If you have multiple unit operations with very similar tasks, there’s no need to reinvent the wheel.  You can clone tasks from one unit operation to another easily.  This makes serving 10s, 100s, or even 1000s of operations much easier to set up.

Customer Add-Ins

Many of Intellect’s tasks are “add-in-able”.  By that, we mean you can add in your own algorithms and run them in Intellect.  Create a DLL with a defined API and tell an Add-In task to run like any other in Intellect.

See Problems Before They Occur

With predictive analytics you can estimate the likelihood of problems, product performance, process performance, faults, or other adverse events soon enough to take action to prevent them.

Gain Insights into Performance

Through data modeling and "sensitivity analysis," you can quickly identify the key drivers that impact your product performance and KPIs most from a qualitative and quantitative standpoint.

Manage and Control Complex Processes

Use Model-predictive control (MPC), a powerful tool for controlling complex processes, to maintain precision in your process performance and reach new levels of excellence. IntelliDynamics' MPC enables you to determine multiple setpoints in realtime.