Digital twins for the process industry
Live models of real production lines. I combine physics-based simulation with sensor data so operators can see what is happening inside a die, a mould or a casting line, and act on it.
Research Scientist at IFE, Kjeller, Norway
Sayan
Adhikari
I build simulation and data tools that let industry see inside its own processes, from finite element models of aluminium casting to agentic AI that sets up the simulations for you.
electrons ions Live particle simulation in a magnetic field. Your pointer is a biased probe; click to flip its polarity.
From plasma sheaths to production lines
I am a computational physicist in the Institute for Energy Technology's Computational Materials Processing group near Lillestrøm, Norway. I develop mathematical models and the software around them: finite element and multiphysics solvers, the data infrastructure that feeds them live measurements, and the AI layer that makes them usable by people on the factory floor.
My current work is with manufacturers in aluminium, battery and solar production, where I build digital twins of casting and extrusion processes and explore how agentic AI can automate simulation workflows. I take an active part in shaping research projects with industry, writing proposals for the Research Council of Norway and Horizon Europe.
Before IFE I spent a decade in plasma physics. I earned my Ph.D. at the Centre of Plasma Physics in India modelling magnetised plasma sheaths, then joined the University of Oslo to simulate Langmuir probes and spacecraft charging for missions such as the Emirates Lunar Mission's Rashid rover.
- Experience
- 12+ years in numerical modelling and code development
- Publications
- 46 papers, preprints and proceedings; cited 285 times, h-index 10 (Google Scholar, October 2026)
- Programming
- Python, C/C++, Fortran, MATLAB, Java
What I work on
Research in detailMultiphysics and finite element modelling
Thermo-mechanical and flow models of material processing, built to run inside industrial projects rather than only in papers.
Agentic AI for simulation
Teams of LLM agents that read a problem in plain language, build the mesh, run the solver and check the result, backed by a knowledge graph of the physics.
Plasma and space physics
Where I started: kinetic and fluid simulation of plasma sheaths, probes and spacecraft charging, from laboratory devices to the ionosphere and the Moon.
Projects
-
2026 – 2030
SAGA, Smart Aluminium for Grid and Energy Applications
Innovation project for the industrial sector (IPN) led by Hydro Aluminium, funded by the Research Council of Norway. Aluminium wire rod for the power grid with more post-consumer scrap and stable electrical properties. I work on casting and rolling simulation, physics-informed surrogates and agentic AI for the production line.
-
2026
PDE-Agents, LLM agents that run real simulations
A team of LLM agents that turns a plain-language description of a physics problem into a verified finite element simulation. It runs on local open models, so geometry and material data never leave the site, and every reasoning step is logged for audit. It now drives our aluminium casting code and OpenFOAM cases such as battery anode slurry mixing. Preprint on arXiv, DigiPro webinar, 2026.
-
2022 – present
SUCCEED and FLEX4FACT, digital twins for polymer extrusion
I lead IFE's part of SUCCEED, an IPN project with Inaventa Solar, Omipa, Solvay and Fraunhofer ISE that builds a real-time digital twin of the extrusion line for high-performance polymer solar collectors. The same line is a use case in the Horizon Europe project FLEX4FACT, where we combine CFD and thermal models of the die with live process data and machine learning.
-
2020 – 2022
Langmuir probe on the Rashid lunar rover
At the University of Oslo I modelled the plasma environment of the probe instrument built for the Emirates Lunar Mission, alongside probe studies for NanoMagSat and Daedalus within the 4DSpace initiative.
Recent publications
All 46 publications-
2026
-
2026
Toward a digital twin: Extrusion die for manufacturing polyphenylene sulfide solar thermal collectors
Polymer Engineering & Science 66 (3), 2130–2140
-
2025
The Langmuir probe instrument on board the Rashid-1 rover of the Emirates Lunar Mission
Space Science Reviews 221 (3), 34
-
2025
Enhancing extrusion efficiency: Development of a digital twin for glass reinforced polymer processes using machine learning and real-time data integration
TMS Annual Meeting & Exhibition, 620–633
Path so far
Full CV-
2022 – present
Research Scientist, Institute for Energy Technology (IFE), Kjeller, Norway
-
2022
Researcher, University of Oslo, Norway
-
2020 – 2022
Postdoctoral Fellow, University of Oslo, Norway
-
2017 – 2022
Co-founder, CTO and Creative Director, Good Morning Science, India
-
2018 – 2020
Postdoctoral Fellow, Centre of Plasma Physics, Institute for Plasma Research, Assam, India
-
2012 – 2017
Junior and Senior Research Fellow, Centre of Plasma Physics, Institute for Plasma Research, Assam, India
-
2019
Ph.D. in Plasma Physics, Gauhati University, India
Let's work together
I am happy to talk about industrial modelling projects, digital twins, agentic AI for simulation, or consortia for Norwegian and European calls.