Research & Insights
The thinking behind the world model for energy.
Customer stories, articles, talks and technical studies from the team building Orbital: plant data, physics, and what it takes to run AI in daily operations.
Featured customer story
Indorama Ventures: verified by the engineers who run the plant.
At the F5 PO/MTBE unit in Port Neches, Texas, the engineers tested what Orbital understood about their plant, section by section. They found reasons to expand it.
Analyst view · ARC Advisory Group
ARC View on the Applied Computing Demo Day 2026.
An independent ARC Insight report on our 2026 Demo Day. ARC places Applied Computing in an emerging class of industrial AI vendors building foundation models for physical systems, rather than enterprise productivity tools.
Jan Burian, Research Director, ARC Advisory Group · 27 August 2026
The work
Filed by the question it answers.
Proof on the plant
What happens when the engineers who run a unit test Orbital.
Inside Orbital
One in-depth article per product: the problem it solves, how it works and where your engineers stay in control.
Orbital Plant Intelligence
Plant Intelligence: operate and optimise the plant, with engineers in charge
Most plants already run well. The opportunity is the last few per cent of yield, efficiency and reliability. Orbital Plant Intelligence catches problems early, knows product quality between lab samples, finds today's best operating point and tests changes first, then lets engineers build their own tools by describing them.
Orbital Plant Design
Designing and changing plants: Orbital Plant Design for capital projects
Orbital Plant Design, in limited early access, brings the world model to capital projects and plant design: through basic, front-end and detailed engineering on new plants, and for capital projects on the plants Orbital already helps run, starting from drawings kept in step with the plant.
Orbital Data Connect
Before the AI: why plant data is the first problem to solve
Most plant data never informs a decision. Orbital Data Connect brings historian, lab, document and maintenance data into one governed place, so AI can be trusted with it from day one.
Research: the world model for energy
A four-part series on how a world model learns a plant, why physics belongs inside it, and how Orbital turns a question into a decision.
Research · Part 1 of 4
Why large language models fall short in process plants
A plant describes itself in sensor data, documents and drawings that never share a language. A language model on its own gives fluent answers, not decisions. This is why we split the problem in two.
Research · Part 2 of 4
What is a world model for energy? Inside the Orbital world model
Understanding a plant takes more than its data. The Orbital world model reads a plant through its signals, its physics and its documents, so every answer about a unit tells one consistent story.
Research · Part 3 of 4
Physics-informed AI: why physics belongs inside the model
A model that has only seen data knows which answers look like the past. It does not know which answers the plant can physically deliver. Plausible is not feasible, and physics is what closes the gap.
Research · Part 4 of 4
From question to decision: a world model on a catalytic reformer
The Orbital world model is the foundation. Orbital is the system that puts it to work. Here is one operating question on a hypothetical catalytic reformer, followed from start to finish.
Scaling AI in industry
What analysts, broadcasters and conference stages say about taking AI from pilot to operations.
Analyst report
ARC View on the Applied Computing Demo Day 2026
ARC Advisory Group’s independent read on Orbital, physics-grounded AI and where industrial foundation models fit.
Podcast
How can AI be scaled effectively in chemical manufacturing?
A conversation on the S&P Global Chemical Week podcast about moving AI from pilots to operations.
Keynote
AI and the energy security problem
A keynote at Web Summit Qatar on the challenges facing the energy industry.
Interview
Why AI pilots fail to scale
ET EnergyWorld at India Energy Week 2026, on India’s rising energy demand and taking AI from pilots to production.
Perspectives from our leaders
How the people building Orbital see the industry and where it is heading.
From the podcast
Dan Jeavons on the Future of Energy Intelligence
Only 8% of energy data is used today. Dan Jeavons explains how Orbital, Applied Computing’s foundation model for energy, is unlocking the rest, helping redefine the future of energy intelligence in the AI age.
Perspective
Reflections from 35 Years in Manufacturing: Why The Future of Energy Operations Lies in System Intelligence
For most of my 35-year career in manufacturing, I have seen the industry evolve dramatically but struggle with one persistent challenge.
Perspective
The Climate Case for Drilling: Why Domestic Oil and Gas Serves as a Responsible Bridge to Long-term Sustainability
A deep dive into the pragmatic case for continued domestic oil and gas production that breaks down how responsible drilling can reduce emissions, protect sovereignty, and serve as the bridge to a renewable future.
Technical studies
The studies behind Orbital.
Each study looks at a specific prediction, optimisation or early-warning problem. Every one is labelled with the setting it comes from.
Carbon capture
Better prediction of CO₂ concentration
Benchmarked on a research-scale carbon capture pilot plant, against the plant’s state-of-the-art model.
Read the study
Distillation
Less energy in a simulated distillation process
A DWSIM simulation of the alkylation of 1-butene to octane and dodecane, holding product quality.
Read the study
Gas processing
Higher yield purity in simulated natural gas processing
A DWSIM simulation of a gas processing plant producing LPG, pentane and lean gas.
Read the study
Refining
A year in savings identified in a simulated hydrotreater
A DWSIM simulation of a refinery hydrotreater, reducing energy use while holding product quality.
Read the study
Asset reliability
Earlier failure predictions than other models
Pipeline failure prediction up to 96 hours ahead, tested in a simulated environment.
Read the study
More articles
From the archive.
Earlier articles from the team on AI in energy, physics-grounded AI, change management and the energy transition.
Applied Computing
Unlocking the 92% of Untapped Data: How Orbital is Defining the Future of AI in Energy
AI in Energy Transformation
Game-Changing Intelligence: How AI Use Cases in Sports Can Apply to the Energy Industry
AI Use Case
Preventing $1–2 Million in Downtime Losses with AI-Powered Predictive Maintenance
Change Management
The Crucial Role of the Managerial Middle in AI Change Management
AI in Energy Transformation
AI and Offshore Wind: Building Resilience in the World’s Harshest Environments
Physics-Grounded AI
Physics-Grounded AI is Rewriting the Rules of Fluid Dynamics
Change Management
The Human Side of AI: Rethinking Change Management in an Age of Intelligent Machines
Physics-Grounded AI
Building the Future of Intelligent Manufacturing with Physics-Grounded AI
Applied Computing
Applied Computing bolsters its senior team
ESG & Transition
How AI-Powered HVAC Systems Deliver 8% Energy Savings for Sustainable Buildings
From our advisors
Experts speak.
What impresses me about Orbital is the trust it creates. Insights aren’t just clever, they’re grounded in fundamentals, contextual and ready for real operational decision-making.
