From two decades of research to the furnace floor.
Smart Ironmaking Computing (SIC™) builds physics-constrained AI for ironmaking. Our models come from more than 20 years of blast-furnace research and over 200 peer-reviewed publications.
Mission
Make every blast furnace transparent enough to decarbonise with confidence.
Steelmakers know what they need to change. What they lack is a trustworthy view of what each change will do inside the furnace. We build that view.
Physics first
AI that respects heat, mass and momentum balance — or it isn't fit for a furnace.
Engineers, not slogans
Every recommendation comes with numbers your team can check.
Confidential by design
Plant data stays protected, encrypted and under your control.
What we believe
The unmeasurable is not the unknowable.
No sensor today can reach the cohesive zone, gas flow or temperature field inside a blast furnace. But they obey heat, mass and momentum balance — so they can be computed. That is why SIC™ and Agent X exist.
Four ideas behind our method
Hover over or tap a poster to see the method behind it.
Computational fluid dynamics
Everything flows
Gas, burden and molten iron never stop moving inside a furnace. We solve those flows with CFD, and read the state of the furnace from how they move — the line long attributed to Heraclitus, put to work.
Everything flows
Finite-volume meshing
Divide each difficulty
Descartes divided every difficulty into parts; the Great Learning asks us to investigate things to extend knowledge. We divide the furnace into many thousands of cells, solve mass, momentum and heat balance in each, and assemble the whole.
Divide each difficulty
Physics-generated data
From one law, all things
Laozi wrote that the Tao produced One, and in time all things. In Agent X the Tao is physical law: from one set of conservation equations we generate in-furnace data across many operating conditions — the case database the agent answers from.
From one law, all things
Mechanism models + plant data
Learning and thought
Learning without thought is labour lost; thought without learning is perilous. Data without physics cannot explain, and physics without data drifts from the plant. Agent X keeps each in check with the other.
Learning and thought
Leadership
The team
Prof. Yansong Shen FTSE
Co-founder & Chief Scientist
Fellow of the Australian Academy of Technological Sciences and Engineering. He led the team that built the first industrial-scale, three-dimensional transient model of a whole blast furnace with resolved ore and coke layers, coupling gas, liquid and solid phases with chemical reactions — and a multi-scale simulation platform from particle to whole-furnace scale. His Renewable Injections–Sustainable Burdens (RISB) process for near-term blast furnace decarbonisation is supported by the Australian Renewable Energy Agency (ARENA). As Chief Scientist, he sets SIC™'s scientific direction.
Specialist in blast furnace modelling and industrial simulation, with project experience across many steelmakers — he turns plant needs into technology that can be deployed. Leads the company and the development of SIC™ BF Smart Agent X.
Core developer of the world's first three-dimensional transient blast furnace model, coupling multiphase flow, chemical reactions and heat transfer. An expert in ironmaking processes and CFD, he leads SIC™'s technology and the engineering of Agent X.
Working with a leading blast-furnace research team
SIC™ works with SCoPE Lab for Green Metal, a leading research team in blast furnace modelling and green metal production, to keep Agent X at the front of ironmaking science.
Work with us
We welcome enquiries from engineers, researchers and students interested in ironmaking, CFD and industrial AI.
