26.02.2026

Meghana Sudarshan has begun work in one of two new positions funded through the Research Council of Norway’s scheme for recruiting talented researchers. She will build an open‑source machine learning toolbox that will help researchers and industry understand batteries faster and with greater precision.

FME BATTERY has secured two of these estemeed positions, designed to strengthen Norwegian research fields by attracting highly skilled early‑career scientists. While the second researcher is expected to start at UiA later this year, Sudarshan has already taken up her role.

“I'm really proud to receive funding through the Recruitment of Talented Researchers scheme. It has been great to welcome Meghana onboard at IFE, and I'm excited about the new perspectives and added focus that her work will bring to FME BATTERY,”
- Hanne Flåten Andersen, Director of FME BATTERY

Understanding Battery Behaviour Through AI

Sudarshan’s research centres on using artificial intelligence to better understand how batteries behave. She recently completed her PhD in applied machine learning for batteries at Purdue University in Indiana, USA.

Originally from India, she has travelled far, both geographically and academically, to join IFE’s research community at Kjeller, just outside Oslo.

Her experience includes working with a range of industrial partners, includingTesla. Those collaborations highlighted the gap she hopes to help close.

“I realised industries face specific practical hurdles that academia hasn’t always addressed. I love that IFE and FME BATTERY is right in the middle of that, to bridge the gap and keep industry and academia in sync.”

Building an Open‑Source Toolbox for the Battery Community

A key motivation behind the project is the growing need for large, well‑structured datasets to understand how and why batteries degrade.

“To understand how long batteries last, you need a lot of data,” says Sudarshan.
“Everything is battery‑powered here in Norway! There is an abundance of battery data, but it is not well used yet. A toolbox like this would be helpful to, for example, detect the cause if something goes wrong with a battery.”

Over the next three years, Sudarshan and her colleagues will develop an open‑source modelling toolbox accessible to both researchers and industry.

The planned ML‑Batt‑TOOL will consist of three key components:

• Battery prompts and diagnostic signals: a structured way to detect and classify battery behaviour.
• Virtual battery profiles for edge-cases: supporting model training when realworld data is limited or fragmented.
• Predictive simulation tool: enabling users to model and predict battery behaviour under different conditions.

“The MLBattTOOL will be an opensource AI toolbox designed to help researchers and industry work faster and more accurately,” Sudarshan explains.

Sudarshan’s work is directly affiliated with work package 6 in our centre: Digitalisation.

• Read more about Meghana Sudarshan at her website: https://msudars.github.io/